DEV Community: Oleksyj Mokrousov

Keyword vs Google Taxonomy: Fix Noisy Vertical News Alerts

Oleksyj Mokrousov — Sat, 25 Jul 2026 10:23:06 +0000

Your industry alert is still a string match

A “banking desk” built on q = bank will happily ingest river banks, blood banks, data banks, and robbery blotter. A “rock feed” on q = rock will mix geology, wrestling, and cocktail recipes with actual Rock Music coverage. The channel looks productive; the product is not on-vertical.

This page is a bakeoff diagnostic for noisy keyword desks. It is not the full Google taxonomy filter cookbook (exact path vs prefixTaxonomy, Stories, classify). That lives on Perigon’s News API filter by Google taxonomy guide - link out when you are ready to wire the params.

Why do vertical keywords inflate alerts?
How large are the gaps for bank, fund, and rock?
When should you use taxonomy vs Perigon category vs topic?
What bakeoff should you run before shipping a desk?

Why vertical keywords inflate alerts

Keywords match tokens. Vertical desks need aboutness. Google Content Category paths (V2) on Perigon articles describe industry and content family - Banking under Finance, Rock Music under Arts & Entertainment - so a filter can mean “this kind of story,” not “this substring appeared.”

Habit	What you get
q = bank / banking	Central banks, rivers, features named “bank,” crime, metaphors
q = fund	Mutual funds plus “fund a project,” fundraisers, verb uses
q = rock	Music plus geology, sports nicknames, food, idioms
Taxonomy path for the vertical	Articles classified into that Google hierarchy

You still stack dates, sources, and companies. The fix is choosing the right aboutness filter first.

Live bakeoff: bank, fund, rock (30 days)

Articles counts for pubDate 2026-06-25 through 2026-07-25 (round in product copy; huge open queries are order-of-magnitude):

Desk intent	Keyword filter	About how many	Taxonomy / structured filter	About how many
Banking vertical	q = bank	~988,000	taxonomy = /Finance/Banking/Other	~363,000
Banking (wordier keyword)	q = banking	~259,000	prefixTaxonomy = /Finance/Banking	~363,000
Funds vertical	q = fund	~695,000	taxonomy = /Finance/Investing/Funds	~398,000
Rock music vertical	q = rock	~341,000	taxonomy = /Arts & Entertainment/Music & Audio/Rock Music	~210,000

Read the table carefully. Taxonomy is not always the smallest number - q = banking (~259k) is below Banking taxonomy (~363k) because many banking stories never use the token “banking.” The failure mode is different: keywords miss on-vertical pieces and include off-vertical tokens. Taxonomy stays on the Google Banking branch either way.

Keyword bank (~988k) is still ~2.7× the Banking taxonomy slice - that gap is mostly noise and adjacent senses. Keyword fund (~695k) vs Funds path (~398k) shows the same pattern. Keyword rock (~341k) vs Rock Music (~210k) cuts a large geology/idiom/sports tail while keeping tour and album coverage the music desk wants.

For a wider finance firehose, prefixTaxonomy = /Finance was about 3.4M in the same window - useful as a branch monitor, not a tight desk. Pair a narrower leaf path when alerts must stay readable.

Taxonomy vs category vs topic (pick one job)

Perigon exposes three related dimensions. Mixing them up recreates keyword chaos with nicer names.

Filter	What it is	Use when
taxonomy / prefixTaxonomy	Google Content Category paths	Industry / content-family desks (Banking, Rock Music, Health Conditions)
category	Perigon broad themes (Finance, Tech, Health, …)	Fast dashboard buckets; e.g. category = Finance was ~3.5M in this window
topic	Granular Perigon topics	Storyline types - e.g. topic = Cryptocurrency ~187k vs q = cryptocurrency ~89k (topic can out-recall a thin keyword)

Rule of thumb: Google path when the desk is an industry tree; Perigon category for coarse theme chrome; topic when you need a named storyline slice. Stack carefully - do not OR three overlapping definitions of “finance” and call it precision. Deep param behavior stays on the taxonomy filter guide linked above.

Bakeoff checklist before you ship a vertical alert

Write the desk intent in one sentence (“Rock Music coverage,” not “anything with rock”).
Run q = token and the matching Google taxonomy (or prefixTaxonomy) for the same from/to window.
Spot-check twenty titles from each list - count off-vertical hits and obvious misses.
If the keyword is quieter but incomplete (like q = banking vs Banking path), prefer taxonomy and accept the larger on-vertical set.
Only then add source packs, companies, or labels - aboutness first, polish second.

Key takeaways

Vertical q filters are string matches; taxonomy paths are aboutness filters for industry desks.
Live gaps matter: bank ~988k vs Banking taxonomy ~363k; rock ~341k vs Rock Music ~210k (30 days ending 2026-07-25).
Quieter keywords are not always better - banking can under-recall vs the Banking path.
Use taxonomy for industry trees, category for broad themes, topic for granular storylines.

Place-Name Keywords vs Geo Filters for Local News

Oleksyj Mokrousov — Sat, 25 Jul 2026 10:12:39 +0000

Your “local” alert is probably a keyword in disguise

Many desks build “Chicago news” or “Portland coverage” by stuffing the city into a keyword field. That habit creates false confidence: the feed looks busy, the Slack channel never sleeps, and half the hits are the wrong market, a person named Austin, or a national wire that mentioned the city once.

This page is a diagnostic, not a full geography-param cookbook. It shows five failure modes with live volume gaps, then points you to structured filters. For the complete local news API surface (county, radius, Stories, Sources), use Perigon’s Local news API search guide.

What goes wrong when the city is only a keyword?
How big is the gap for ambiguous markets (Portland, Austin, Springfield)?
How do event-city and newsroom-city filters differ (Chicago)?
What should you change in the alert job tomorrow?

Try it: Articles sandbox · Docs: Local news · Plans: API pricing

Failure mode 1: Ambiguous city names (the Portland problem)

Portland is not one market. A keyword match cannot tell Oregon from Maine - or from cement, travel, and sports noise that shares the string.

Live Articles counts for pubDate 2026-06-25 through 2026-07-25:

Filter	About how many articles
q = Portland	~67,000
city = Portland + state = OR	~6,200
city = Portland + state = ME	~1,300

Keyword Portland is roughly 10× Oregon event-scoped volume and 50× Maine. If your “Portland desk” used q, most of the firehose was not your DMA.

Failure mode 2: Person and brand collisions (Austin)

Austin is a capital city and a common given name. Keyword search pulls both.

Filter	About how many articles
q = Austin	~143,000
city = Austin + state = TX	~11,000

Texas event geography is about 13× quieter than the keyword - still busy, but no longer every headline that mentions someone named Austin.

Failure mode 3: Many towns, one string (Springfield)

Springfield is a classic trap: Illinois capital, plus cities in other states, plus fiction and sports nicknames.

Filter	About how many articles
q = Springfield	~24,000
city = Springfield + state = IL	~2,200

If the job is Springfield, Illinois government or plant news, the keyword feed is mostly other Springfields and incidental mentions.

Failure mode 4: Event city vs newsroom city (Chicago)

Even when the place string is unique enough, where the story happened and where the publisher sits are different questions. Keyword Chicago mixes both - and everything else.

Filter	About how many articles
q = Chicago	~321,000
city = Chicago + state = IL (event / story place)	~60,000
sourceCity = Chicago + sourceState = IL (publisher based there)	~144,000

Event-scoped Chicago is far below the keyword firehose. Newsroom-scoped Chicago is larger than event-scoped - national and metro outlets headquartered in Chicago cover many non-Chicago events. Pick the axis on purpose: city council and weather emergencies → event city / state; hyperlocal reader apps and DMA ad products → sourceCity / sourceState. Combining both is a third product (events in a county from in-state publishers) - details in the local news guide linked above and location & geo docs.

Failure mode 5: Treating volume as quality

A keyword city alert “working” usually means it is loud. Use a one-week bakeoff before you ship:

Run q = CityName and city + state side by side for the same from/to window.
Spot-check twenty titles from each - count wrong-market and name-collision hits.
If you need outlets in the market rather than events there, swap in sourceCity / sourceState (or discover publishers on the Sources API first).
Only then add topic, company, or radius - geography first, keywords second.

Radius feeds (lat / lon / maxDistance) belong when the user or plant is a point on a map, not a city string - again owned by the full local guide, not this diagnostic.

What to change in the alert job tomorrow

If your job looks like…	Stop doing	Start doing
“Portland OR desk” via q	Keyword Portland	city + state = OR
“Austin CI” via q	Keyword Austin	city + state = TX (+ company filters)
“Springfield IL plant”	Keyword Springfield	city + state = IL (or county)
“Chicago hyperlocal app”	Keyword Chicago	sourceCity / sourceState, or Sources then allowlist
“Stories in the metro”	Article keyword flood	Stories with city / state (see local guide)

Key takeaways

Place-name keywords are not local news monitoring - they are string matches with market collisions.
Live gaps are large (e.g. Portland q ~67k vs OR ~6.2k vs ME ~1.3k; Austin q ~143k vs TX ~11k in 30 days ending 2026-07-25).
Chicago shows event city (~60k) and sourceCity (~144k) answer different product questions - neither equals keyword Chicago (~321k).
Fix the geography axis first; leave the full param and radius cookbook to Perigon’s Local news API search guide.

Get started: Articles sandbox · Local news docs · Local news API search guide · API pricing

News Integration: RSS Polling vs Structured Feeds

Oleksyj Mokrousov — Sat, 25 Jul 2026 09:46:35 +0000

Product news features hit a wall that feed readers never see

Shipping a news integration - a Slack alert bot, a customer dashboard tile, an AI agent that cites today’s coverage - feels easy if you start with RSS polling. Grab a few XML feed URLs, cron every 15 minutes, parse title/link/date, done. That pattern works for humans in a feed reader. It falls apart the moment the same stream has to power a product: dedupe across outlets, filter by company or person, push events into your system, or keep latency predictable under load.

This piece is for builders. If you only need a personal reader decision (when Google News / publisher RSS is enough vs curated trackers), start with Perigon’s Google RSS vs data feeds guide instead - that page owns the human-monitor angle. Here the job is different: news API for products, production news pipelines, and why XML news feeds stop being an architecture.

The sections below answer:

What does RSS polling look like inside an app stack?
Where do XML news feeds break product pipelines?
What does a structured news integration look like instead?
When should you keep RSS vs move to feeds, Signals, or the API?
How do live coverage volumes change the design?

Try queries: Articles sandbox · Push events: Signals webhooks · Plans: API pricing

What RSS polling looks like in an app stack

RSS polling is a client-side loop: store feed URLs, HTTP GET on a schedule, parse channel items, upsert by link or GUID, notify users. Many teams also scrape Google News–style query feeds the same way. The contract is thin - usually title, link, summary, published date - and each URL is its own silo.

In a product that means:

Layer	Typical RSS approach	Product friction
Ingest	Cron + XML parse	Poll windows miss bursts; every publisher is a custom URL
Identity	Link / GUID	Same story across 20 outlets → 20 rows
Filter	Keyword rules in your code	No stable company or person IDs
Delivery	Your own queue after poll	You own retries, backlog, and quiet hours
Scale	Add more feed URLs	Coverage sprawl without shared schema

None of that is wrong for a hobby bot. It is a weak foundation for a structured news feed your customers rely on.

Where XML news feeds break product pipelines

Four failure modes show up once the feature leaves a side project.

Volume without structure. In the Perigon index for pubDate 2026-06-25 through 2026-07-25, a plain q of cybersecurity matched about 138,000 articles. OpenAI as a keyword hit about 167,000; resolving the same company via companyId still returned about 128,000 articles but only about 402 Stories clusters. An RSS inbox treats every reprint as a new item. A product that alerts on “OpenAI news” without clustering or entity filters drowns users.

Keyword ≠ entity. The same 30-day window: ChatGPT as keyword ~80,000 articles; ChatGPT plus the OpenAI companyId ~33,000. Sam Altman via person Wikidata Q7407093 ~17,500 articles / 61 Stories. RSS has no first-class companyId or personWikidataId - you reinvent that with brittle string rules.

Ticker and brand noise. companySymbol NVDA alone was about 178,000 articles in that window; keyword NVIDIA about 163,000. High-volume names need allowlists, labels, and source packs - not another XML URL in the poller.

Push vs pull. Products want webhook news alerts into Slack, a CRM, or an internal bus. RSS is pull-only. You own the scheduler and the lag. Structured monitors that emit JSON events fit product SLAs better than “we polled 40 feeds at :00 and :15.”

What a structured news integration looks like

A production news integration usually separates three jobs RSS tries to do alone:

Retrieve - queryable articles and stories with filters (company, person, source group, dates), not only subscribed XML URLs.
Monitor - a durable watch (Signal / curated feed) for the beat your product cares about.
Deliver - email, connectors, or webhooks into your systems when something matches.

On Perigon that maps to the Articles / Stories APIs for retrieval, Signals and Discover feeds for ongoing trackers, and Signals webhooks for push into your stack. The human-facing “is RSS enough?” decision stays on the RSS vs data feeds guide linked above; this page assumes you already know you are building software.

Capability sketch for builders:

Need in the product	RSS polling	Structured feed / API path
Stable company or person filter	DIY keywords	Entity IDs on search and monitors
One narrative, many reprints	Manual dedupe	Stories / clusters
Real-time push to your app	Build it	Webhooks and connectors
Shared schema for UI and agents	Title + link	Structured article fields
Desk theme without URL sprawl	Folder of feeds	Curated / Signal-backed feeds

For ad-hoc query design (Boolean q, sources, pagination), use the practical news API guide - this post does not re-teach those params.

When to keep RSS vs when to move

Keep RSS polling when the feature is a personal or internal reader, the outlet list is short and stable, and nobody outside that reader needs a shared schema or push SLA.

Move to structured feeds / Signals when the same beat must power a customer-facing surface, a team channel, or an automation - and you need multi-source coverage without babysitting feed URLs.

Move to the News API when the product needs programmable search, custom filters, historical windows, or agent tooling on top of the index - not only a prebuilt tracker.

A practical migration for an app team:

Inventory which RSS URLs actually drive product behavior; drop vanity polls.
Map each job: awareness UI vs alert vs AI grounding.
Parallel-run: keep the poller for a sprint while a Signal-backed feed or API query covers the same beat; compare misses, duplicates, and latency.
Graduate push paths to webhooks; keep RSS only where a human still wants a classic reader.

Key takeaways

RSS polling is a fine reader pattern and a fragile news integration for products.
Live volumes (e.g. ~128k OpenAI-company articles vs ~402 Stories in 30 days ending 2026-07-25) show why apps need entities and clusters, not more XML URLs.
Split retrieve / monitor / deliver: API + Signals/feeds + webhooks beat a single cron parser.
For “is RSS enough for a human desk?” use the Perigon RSS vs data feeds guide; for shipping product pipelines, design for structured feeds and push delivery.

Get started: Articles sandbox · Discover feeds · Google RSS vs Perigon data feeds · API pricing

44 Best Manufacturing News Sources in 2026

Oleksyj Mokrousov — Sat, 25 Jul 2026 09:37:15 +0000

AI top-8 industrial source lists miss most of the beat

Teams buying manufacturing news, factory automation news, and industrial automation news feeds often start by asking an LLM for the “best” industrial trade publications. That is a weak foundation for production monitoring. GPT, Claude, and Gemini each invent a short industry news sources roster - they disagree on most names, skip whole lanes (supply chain trade desks, UK construction press, additive manufacturing, pumps and reliability), and still feel complete because eight domains fit in a chat reply.

Do not trust that allowlist. A brainstorm is not a curated B2B manufacturing media pack. Perigon’s industrialTradePublications source group (ID 725, 44 domains) is built for the full industrial picture: manufacturing leadership, plant engineering, controls, machinery, robotics, process ops, and adjacent supply chain trade news - then applied as one reusable sourceGroup on search.

Rather than maintaining dozens of publisher filters yourself, Perigon groups related publications into reusable source groups. Instead of listing every manufacturing publication in each API request, you can reference a maintained group that evolves over time.

What fuller vertical manufacturing news coverage looks like in industrialTradePublications source group?

How to monitor machinery news, plant ops, and automation with that source group

How to filter manufacturing news by source, company, topic, category, location, and date

Try it live: Sandbox · Docs: Sources and source groups · Plans: API pricing

Other vertical packs (different SERPs): healthcare trade pubs · top25finance. Creating private forks: Sources and source groups - this page applies the public industrial pack only.

Why you should not trust AI for industrial trade publication allowlists

Prompt three models for top industrial trade publications with domains. Across ChatGPT, Claude, and Gemini:

Only two outlets appear on all three lists: IndustryWeek (industryweek.com) and Automation World (automationworld.com).

Everything else is a bias: Claude skews assembly / ENR construction / American Machinist; Gemini pulls Manufacturing Dive, power, chemicals, and packaging titles; GPT favors machine-shop and UK engineering picks.

Overlap is thin. Two-model agreement reaches a few more (manufacturing.net, plantengineering.com, controleng.com) - still nothing like a full niche industrial sources set.

What AI does	What breaks in production
Returns a neat “top 8”	Misses 30+ trade publication news domains you may already need
Sounds authoritative	Different models → different desks → different alert quality
Mixes brainstorm + “also include” extras	No stable name your Articles / Stories jobs can keep forever

Claude even warns that “industrial” shifts by sub-sector (aerospace, materials handling, oil and gas). That is the point: an LLM answer is a partial map. For plant engineering news, factory automation news, and industryweek-class leadership coverage plus the rest of the floor, you need a maintained sourceGroup - not a chat paste.

Quick split (domains once):

Claude-only top-8 → assemblymag.com, machinedesign.com, ien.com, enr.com, americanmachinist.com.

Gemini → manufacturingdive.com, packworld.com, powermag.com, chemengonline.com.

GPT-only → mmsonline.com, processingmagazine.com, theengineer.co.uk.

Treat those as research signals that AI is incomplete - not as three competing truths.

The fuller picture: industrialTradePublications group with 44 domains

Field	Value
Name	industrialTradePublications
Display name	Industrial Trade Publications
Source group ID	725
Size	44 domains
Scope	Trade sources covering manufacturing, automation, machinery, and industrial operations

Pass sourceGroup=industrialTradePublications on Articles (/v1/all), Headlines, or Stories. Inspect members via Sources with the same group name - without dumping every domain into this article twice.

What Perigon keeps that AI top-8 lists drop

Durable AI shorts that do belong in a serious pack are already here (IndustryWeek, Automation World, Manufacturing.net, Plant Engineering, Control Engineering, plus many Claude/Gemini/GPT singles). The fuller picture is the lanes chat lists routinely skip:

Daily manufacturing + supply chain trade news - e.g. Manufacturing Dive and supply-chain trade desks (manufacturingdive.com, supplychaindive.com, supplychaindigital.com, …) beyond any GPT top 8.

Construction / owners / UK build press - more than a single ENR mention (constructionowners.com, constructionnews.co.uk, theconstructionindex.co.uk, …).

Robotics, AMR, digital and additive manufacturing roboticsandautomationnews.com, advancedmanufacturing.org, manufacturingdigital.com, metal-am.com, and peers.

Process, food engineering, pumps, reliability, controls - plant-floor and CPI-adjacent titles (food engineering, pumps and systems, reliable plant, Control Global, chemical engineering) so process manufacturing news is not one magazine guess.

Associations and branded manufacturing media - SME and manufacturing-today / tomorrow style industry news sources models bury under “optional.”

Honest gaps: GPT’s processingmagazine.com and Gemini’s packworld.com are not in this group snapshot. Add them with a private group or source= if mandatory - the public pack is the industrial default, not every URL on the internet.

This is vertical manufacturing news, not healthcare or finance packs (healthcareTradePublications, top25finance).

How to monitor manufacturing and industrial automation news

Scope manufacturing trade publications first, then keywords or companies. Without a publisher gate, live manufacturing OR automation search can surface azonano.com explainers or jobs.magna.com recruiting pages - real pages elsewhere, not industrial trade publications desks. Lock the beat with sourceGroup=industrialTradePublications, then narrow with q (CNC, tariffs, recalls) or entities.

Stack industrial OEMs and automation vendors (resolved IDs)

Combine companyId or companySymbol with the group for OEM / vendor desks:

Company	Prefer	perigon companyId	Symbols to prefer
Rockwell Automation, Inc.	Controls / factory software OEM	0e4fcdf843bf44c69653340aff7729ee	ROK (NYSE)
Caterpillar Inc.	Heavy equipment / plant ops	a6aa46b5af9a45d590694b65741328fc	CAT (NYSE)
Siemens Aktiengesellschaft	Industrial / automation conglomerate	a4aa185497164ee4a72ad9bb8ae2ffe2	SIE.DE (XETRA) or SIEGY (OTC)

Name collisions to skip: Haas / Inductive / Tungsten “Automation” rows; Siemens Energy, Mobility, Healthineers (Healthineers is med-tech, not this pack); credit-union hits under “Caterpillar.” Prefer the parent industrial firm + symbol above.

Pattern: Articles with sourceGroup=industrialTradePublications and companySymbol=ROK (or Rockwell’s companyId) - trade allowlist + vendor, not open web. Company News pillar owns the full entity cookbook.

Example coverage snapshot (minimal, in-group)

Live pull: sourceGroup=industrialTradePublications + companySymbol=ROK - short overviews only (domain + summary; no article URLs):

manufacturing.net - Rockwell threat-detection managed services for industrial customers.

automationworld.com - Rockwell Connected Enterprise / shop-floor connectivity themes.

That is the shape of filter manufacturing news by source after you stop trusting an AI top list.

Use the same sourceGroup on the Stories API when one plant closure, strike, or automation rollout is reprinted across many trade pubs - you track the narrative cluster once instead of every wire copy (see the Stories clustering guide and Stories docs). For a short briefing of that filtered set, pass results into Search Summarizer. Add local / geography filters when the desk cares about a plant city or region, and historical date windows when you need archive depth - keep those recipes in their own guides rather than repeating them here.

Filter manufacturing news by source, company, topic, category, location, and date

Goal	Mechanism	Notes for industrial desks
Publisher allowlist	sourceGroup=industrialTradePublications	Full picture for this vertical
One-off domain	source / excludeSource	Titles AI named that are missing from the pack
Who is mentioned	companyId / companySymbol (ROK, CAT, Siemens AG)	Inside the group
What it is about	topic, category, taxonomy	Taxonomy pillar owns Google paths
Where the outlet sits	Local / source-location	Plant towns, regional construction
When	from / to / addDateFrom	Digests and incremental monitors

When an AI list is a brainstorm vs when you need Perigon’s pack

Situation	Prefer
Ideation slide (“names we recognize”)	AI shortlist - then verify
Production manufacturing news feeds and alerts	industrialTradePublications (725)
Must-have domain absent from the 44	Private group or source=
Publisher metadata	Sources API + group name
Another industry SERP	Healthcare or finance vertical posts

Key takeaways

Do not trust AI top-8 lists for industrial trade publications - models agree on little and miss most of the floor.

industrialTradePublications (725) is the fuller manufacturing trade publications picture: 44 domains across automation, plant engineering, machinery, supply chain, and more.

Filter manufacturing news by source with that sourceGroup; stack OEMs (ROK, CAT, Siemens AG) so job boards and random web hits stay out.

Use LLM lists as research only; use Perigon’s group as the operational allowlist. Keep third-party copy as short overviews - not republished articles.

Get started: Sandbox · Sources and source groups · API reference · API pricing

Monitor NVIDIA News

Oleksyj Mokrousov — Fri, 24 Jul 2026 16:05:25 +0000

Why teams monitor NVIDIA news

NVIDIA news moves across earnings, data-center demand, GPU roadmaps, export-policy spillover, and executive commentary at once. Keyword-only alerts for “NVIDIA” or “NVDA” pull in ticker collisions, supplier noise, and every casual chip mention - while nvidia earnings news, product launches, and leadership statements still get buried.

Teams that need a durable nvidia company news watch usually ask:

How do you monitor NVIDIA for realtime news?

How do you resolve NVIDIA and its leadership before filtering articles?

How do News API entity filters keep NVDA and executive coverage precise?

And we also will answer:

When should you use Signals vs an Articles feed for NVIDIA news search.

The sections below answer each in order. For general company and people entity parameters, see Entities - this page owns the NVIDIA-specific Signals + single-company playbook.

Try it: perigon.io/home · Articles sandbox: sandbox · Entities docs: Entities · Pricing: API pricing

How to monitor NVIDIA news

Perigon makes this straightforward with Signals.

Signals continuously monitor new coverage and deliver updates automatically, so you don't need to rerun the same searches every day.

Open perigon.io/home and describe the watch in plain language. Signals turns that ask into a continuous briefing lane so you are not re-running the same nvidia stock news search every morning.

Use prompts that name the company, the lanes you care about (earnings, GPUs, leadership), and the output shape (bullets, risks first, IR tone). Refine after the first run: tighten to data-center vs gaming, or to executive-only mentions.

Copy-paste Signal prompts

Earnings and guidance watch

Monitor NVIDIA (NVDA) earnings and guidance coverage. Focus on revenue, data-center commentary, gross margin, and outlook. Prefer major financial publishers. Summarize in short bullets for an IR reader; flag contested claims.

GPU and data-center product watch

Track NVIDIA news on GPUs, accelerators, and data-center platforms. Exclude pure consumer gaming rumor roundups unless they affect enterprise supply. Highlight product naming, availability, and partner ecosystem mentions.

Leadership and executive mentions

Watch news that meaningfully mentions Jensen Huang (Wikidata Q305177), NVIDIA founder and CEO. Ignore photo-caption name-checks and unrelated people with similar names. Note interviews, policy comments, and strategy statements that could move investor narrative.

These prompts are evergreen: they describe ongoing beats, not a single week’s headlines.

Resolve NVIDIA and leadership before you filter

Single-company monitoring works best when the org and people are resolved first.

Canonical entities (from Companies + People lookup) in perigon apis

Entity	Use in filters	Resolved value
NVIDIA Corporation	companyId	94bbd0512cd04cc4b7041d78d9d4cf73
NVIDIA (NASDAQ)	companySymbol	NVDA (also listed as NVD.DE on XETRA)
Jensen Huang (CEO / founder)	personWikidataId	Q305177

A name search for “NVIDIA” can also return leveraged/short ETPs (for example GraniteShares or Leverage Shares products).

Prefer NVIDIA Corporation with companyId 94bbd0512cd04cc4b7041d78d9d4cf73 or companySymbol NVDA - not the 3LNV.L / 2NVD.L / 3SNV.L vehicles.

A People search for “Jensen Huang” can surface unrelated people (actors and other Huang/Jensen matches). Keep personWikidataId Q305177 - “American entrepreneur and businessman; founder and CEO of Nvidia” - for nvidia ceo news and jensen huang news.

On the Companies API (/v1/companies/all) and People API (/v1/people/all), re-run lookups when leadership changes; keep Signal prompts and article filters on these IDs. Param reference: Entities.

Articles API recipes for NVDA, earnings, and leadership

On the Articles API (/v1/all), pin the company with companyId set to "94bbd0512cd04cc4b7041d78d9d4cf73" and/or companySymbol set to NVDA. That anchors nvda news and nvidia press coverage to the corporation entity instead of every chip headline or ETP reprint.

Common monitoring scenarios	Start with	Optional add-ons
Core nvidia news feed	companyId or companySymbol=NVDA	sourceGroup=top25finance for IR-grade nvidia investor news
NVIDIA earnings news	Same company pin + q phrases such as earnings, guidance, EPS	Date window around expected print; exclude opinion labels if you want straight news
NVIDIA GPU / data center news	Company pin + q for GPU, accelerator, data center	Topic/category filters when you already use them in product
Leadership / nvidia ceo news	personWikidataId=Q305177	Combine with companyId or NVDA so mentions stay in NVIDIA context
Fresh poll	addDateFrom from last checkpoint	sortBy on add date for ingest-style feeds

Prose example: add a short q for earnings or guidance when you want nvidia earnings news, and poll with addDateFrom so each run only returns what is new since the last check. For executive lanes, add personWikidataId set to Q305177.

For narrative heat around NVIDIA (launches, controversies, multi-outlet spikes), Stories can cluster coverage - use story list filters with the same companyId / NVDA entities. Clustering theory lives in the Stories guide linked in Get started - do not rebuild it here.

When to use Signals vs Articles for NVIDIA

Need	Prefer
Daily briefing without writing queries	Signals on perigon.io/home
Embed nvidia company news in your product	Articles API with companyId 94bbd0512cd04cc4b7041d78d9d4cf73 or NVDA
Leadership-sensitive IR or PR lanes	personWikidataId Q305177 + company pin; optional Signal for narrative
“Is NVIDIA coverage accelerating?”	Stories search with NVIDIA company filters (see Stories clustering guide)
General “how do company filters work?”	Entities docs

Key takeaways

Monitor NVIDIA news with Signals prompts on /home for continuous briefings, then mirror the same beat in the Articles API.

Resolve NVIDIA using the canonical company ID and Jensen Huang using the canonical person ID to avoid ticker and name collisions.

Combine company pins with earnings or GPU q recipes for nvidia earnings news and data-center lanes.

Keep this page single-company; semiconductor-wide and OpenAI-lab watches belong in their own Monitor drafts.

Link Entities for param depth - keep this page NVIDIA-specific.

Get started: perigon.io/home · Entities docs · Stories clustering guide · API pricing

Best Financial News Sources - Filter Market News & Fintech News with top25finance

Oleksyj Mokrousov — Fri, 24 Jul 2026 16:01:10 +0000

Why financial news needs a publisher allowlist

Market news moves faster than most verticals. The same earnings print, Fed decision, or M&A rumor shows up on Bloomberg, CNBC, Yahoo Finance, Seeking Alpha, and a dozen retail investing sites within minutes. Keyword-only news search for "stocks," "banks," or "fintech" also drags in personal-budget blogs, crypto spam, and press-release mills.

If you care about best financial news sources — outlets that actually cover public equities and market movement — lock the publisher set first, then search inside it. That is what Perigon's top25finance source group is for.

Try it live: Sandbox · Sources docs · API pricing

Financial news source group: top25finance

Field	Value
Name	top25finance
Display name	Finance
Source group ID	21
Size	25 domains
Scope	Financial news, market movement, public equities

Pass sourceGroup=top25finance on Articles (/v1/all), Headlines (/v1/headlines), or Stories (/v1/stories/all). Same group name on every endpoint — no 25-domain source = string to maintain.

GET https://api.perigon.io/v1/all?apiKey=YOUR_API_KEY&sourceGroup=top25finance&sortBy=addDate&size=25

Best financial news sources for markets and business press

Domain	Publication	Best for
bloomberg.com	Bloomberg	Global markets, macro, equities data
wsj.com	Wall Street Journal	Corporate investigations, markets
ft.com	Financial Times	International markets and commentary
cnbc.com	CNBC	Live markets, CEO interviews
marketwatch.com	MarketWatch	Stocks, indexes, investing headlines
barrons.com	Barron's	Equities research, weekly outlook
economist.com	The Economist	Macro, policy, global business
fortune.com	Fortune	Corporate strategy, leadership
foxbusiness.com	Fox Business	Markets and economy headlines

Market news, quotes, and retail investing publishers

Domain	Publication	Best for
finance.yahoo.com	Yahoo Finance	Quotes, portfolios, broad market news
nasdaq.com	Nasdaq	Listings, exchange news, market data
investing.com	Investing.com	Indices, calendars, multi-asset coverage
fool.com	Motley Fool	Stock picks, retail investing analysis
seekingalpha.com	Seeking Alpha	Earnings notes, equity commentary
thestreet.com	TheStreet	Trading-oriented market news
investors.com	Investor's Business Daily	Growth stocks, market trends
morningstar.com	Morningstar	Fund and equity research
money.usnews.com	U.S. News Money	Personal finance and investing guides

Fintech news, investing education, and international finance

Domain	Publication	Best for
investopedia.com	Investopedia	Definitions, how-to investing explainers
benzinga.com	Benzinga	Fast movers, trading catalysts
moneymorning.com	Money Morning	Investing ideas and commentary
financialpost.com	Financial Post	Canadian markets
business-standard.com	Business Standard	India business and markets
qz.com	Quartz	Business, tech, global economy
ibtimes.com	International Business Times	International business news

Paywalls: WSJ, FT, Barron's, and friends

Finance is heavier on subscription publishers than many verticals. Inside top25finance, treat these as gated or partially gated:

Domain	Paywall pattern
wsj.com	Strict subscription
ft.com	Strict subscription
barrons.com	Strict subscription
economist.com	Strict subscription
bloomberg.com	Mixed free + premium

Perigon can return articles from some paywalled sources, but that coverage is agreement - and account dependent. It is not turned on for every gated domain by default.

Ask support for FT.com and other paid outlets

If your product needs Financial Times (ft.com), WSJ, Barron's, or similar full-text / expanded paywall coverage.

Contact Perigon support and request paywalled-source access for the domains you need.

Plan integrations assuming FT and peers are opt-in, not automatic.

Open and lightly gated outlets in the group (Yahoo Finance, CNBC, Investing.com, Investopedia, and similar) work under standard API access without that step.

Query recipes for market and fintech feeds

sourceGroup sets the allowlist. Layer search, entities, and time on top for desk-specific filtered news feeds.

Earnings season across the finance bundle

GET https://api.perigon.io/v1/all?apiKey=YOUR_API_KEY&sourceGroup=top25finance&q=earnings&from=2026-06-01&sortBy=date

Fintech keyword slice

https://api.perigon.io/v1/all?apiKey=YOUR_API_KEY&sourceGroup=top25finance&q=fintech&excludeLabel=press%20release&sortBy=date

Desk / product	Tighten with
Breaking markets	source=cnbc.com,bloomberg.com,marketwatch.com
Equity research tone	source=barrons.com,fool.com,seekingalpha.com
Quote / portfolio UIs	source=finance.yahoo.com,nasdaq.com,investing.com
International markets	source=ft.com,economist.com,financialpost.com,business-standard.com
Trading catalysts	source=benzinga.com,thestreet.com
Investor education	source=investopedia.com

Useful companions: addDateFrom for incremental polls, excludeSource to drop one loud domain, excludeLabel=opinion when you want hard news over commentary. Full parameter list: News API reference.

Headlines, stories, and source metadata

Headlines API — same sourceGroup=top25finance when you only need title-level market news surfaces.

Stories API — cluster reprints of one Fed decision or mega-cap earnings print so breaking news monitoring tracks the narrative, not every mirror. Example: topic=Markets with the finance group. See Stories API.

Sources API — inspect the 25 publishers (traffic, geography, paywall flags) before you fork a private group.

Build a narrower custom source list (e.g. US equities only, no education sites) via source groups when the full 25 is too wide.

Compare list: short "best of" vs top25finance

Roundups of best finance websites or best financial news sources usually name eight brands — Bloomberg, CNBC, Yahoo Finance, WSJ, MarketWatch, FT, Investing.com, Barron's — with Reuters, Investopedia, Motley Fool, or TradingView rotated in. Useful starting points; incomplete allowlists.

What short lists usually include	What top25finance adds
1–2 wire / TV desks	CNBC + Fox Business + MarketWatch depth
Yahoo or Investing	Yahoo + Nasdaq + Investing + Morningstar
WSJ / FT sometimes	WSJ + FT + Barron's + Economist (paywall-aware)
One retail voice	Fool, Seeking Alpha, TheStreet, IBD
Rarely education	Investopedia
Rarely non-US	Financial Post, Business Standard, IBT, Quartz

Use the short compare list for product marketing copy if you want. Use top25finance when the financial news provider feed must stay stable across earnings weeks and quiet Fridays alike.

From allowlist to briefing

Typical financial news stack:

Allowlist — sourceGroup=top25finance (request paywall extras if needed).

Signal — q, topic, companyDomain / ticker-oriented terms for the names you trade or cover.

Freshness — addDateFrom + sortBy=addDate for real-time news alerts.

Dedup narratives — Stories API on the same group when reprints explode.

Brief — Search Summarizer over the filtered hit set for a grounded financial news summary.

For structured Finance / Markets category paths, pair the group with taxonomy filters — news taxonomy guide.

FAQ

What does top25finance include?

Twenty-five high-popularity finance domains: flagship markets press, retail investing and quote sites, education, and selected international business outlets. API name: top25finance (ID 21).

How do I get only those publishers in search results?

Send sourceGroup=top25finance on Articles, Headlines, or Stories. That is publisher-level filter news by source for financial news and fintech news without maintaining a domain string yourself.

Can I still use keywords and company filters?

Yes. The group is the outer fence; q, topic, companyDomain, date windows, and excludeSource / excludeLabel narrow inside it.

Are FT.com and WSJ available out of the box?

They appear in the group roster, but paywalled coverage is not guaranteed by default. Ask support to enable FT.com and other subscription publishers for your account when you need them.

Should I use top25finance or paste my own source= list?

Prefer the group for the standard best financial news sources set — Perigon can refresh membership without a client deploy. Paste or fork a private group only when your product needs a deliberate subset (for example, exclude Seeking Alpha, or US-only).

Company News API Guide - Search by Company, Ticker, Domain, and Executive Mentions

Oleksyj Mokrousov — Fri, 24 Jul 2026 15:56:54 +0000

Why company news search needs more than keywords

Building a company news feed on raw text search fails in predictable ways. q=Apple matches fruit and Apple Records. q=Meta hits meta-analyses. Tickers like AI or NOW appear in ordinary English. A CEO's shared surname attaches the wrong person to the result set. Products that need business news data, financial news data, or ticker news require the API to resolve which company or person was linked - not which substring appeared.

Developers evaluating a news API, news search API, or business news provider usually ask:

What is a company news API?

How do you search company mentions and ticker news by domain, symbol, or ID?

Why use company entity filters instead of keyword search alone?

The sections below answer each in order, then give query examples and when to use Articles vs Stories for company-scoped news articles.

Try it live: Test company filters in the sandbox · Read the docs: Entities guide · Compare plans: API pricing and features

What is a company news API?

A company news API is not a separate product SKU. It is a news provider (or financial news provider) that treats organizations and people as entities and lets you filter news articles by those entities.

On Perigon, Articles (/v1/all), Headlines, and Stories (/v1/stories/all) accept company and person parameters. Each hit can include structured fields: company IDs, names, domains, symbols, people, publishers, topics, and dates. Resolve canonical records on /companies and /people once, then pass those IDs into search.

That is the difference between fragile keyword strings and reusable identifiers for b2b data provider news and market data provider news pipelines - the same pattern a financial news or media intelligence product needs behind the UI.

For product-style digests and scheduled briefings across companies or topics, use Perigon's Signals product. This guide is for embedding entity search in your own stack.

How to search news by company domain, ticker, and ID

On the Articles API, company filters match enriched company fields - not a free-text guess.

Parameter	Matches	Typical use
companyDomain	Company web domains (e.g. apple.com)	Brand / website scoped search
companySymbol	Stock tickers (e.g. AAPL)	Ticker news, equities queries
companyId	Perigon company UUIDs	Stable IDs from /companies
companyName	Exact company name	When ID is unknown
personName	Exact person name	Executive mentions by display name
personWikidataId	Wikidata Q-IDs	Disambiguated people from /people

Includes on one field are OR joined. Excludes (excludeCompanyDomain, excludeCompanySymbol, excludeCompanyId, excludePersonName, excludePersonWikidataId) are AND-joined.

Goal	Prefer
Ticker news	companySymbol
Private companies / brand sites	companyDomain or companyId
Executive mentions	personWikidataId first; personName when spelling is canonical

Matched articles return companies and people arrays. Stories return topCompanies and topPeople for narrative-level filters. See the data model for how those objects relate.

Keyword search vs entity filters explained

Keyword q searches title, description, and body. Entity filters ask: was this company or person linked on the article?

Approach	Strength	Weakness
Keyword q only	Flexible phrases ("product recall", "class action")	Ambiguous names, ticker collisions
companyDomain / companySymbol / companyId	Precise org match	Entity must be linked
personName / personWikidataId	Precise people match	Exact name or Wikidata ID required
Keywords plus entities	Event language inside a known org	Best default for litigation, product, crisis

Rival coverage is a clear case: three companyDomain values avoid category-word false hits. A portfolio query ORs companySymbol holdings. An IR or PR query pins a CEO with personWikidataId and adds q for "resignation" or "investigation" when event type matters as much as the entity.

That keeps market intelligence and news filtering result sets auditable: the allowlist defines organizations and people, not shared English tokens.

Company news query examples

How to search rival company domains

Pass one or more rival companyDomain values. Add excludeLabel for press release or opinion when you want earned coverage only. Use addDateFrom when the client polls for new matches since the last checkpoint.

How to search ticker news for a portfolio

OR holdings through companySymbol. Optionally narrow publishers with sourceGroup=top25finance for financial news outlets (domain list and paywalls live in the finance source-group guide). Same symbols on Stories for market-moving narratives instead of every analyst reprint.

And search for such symbols:
PORTFOLIO = ["AAPL", "MSFT", "NVDA"] # companySymbol values

How to search executive mentions

Resolve the person on /people, filter with personWikidataId, and optionally require companyDomain for the employer. Drop collisions with excludePersonName or excludePersonWikidataId.

Product, litigation, and market-moving language

Product / launch: pin companyId or companyDomain, add q for the codename or SKU.

Litigation / risk: same company pin + legal keywords; Stories when one filing reprints widely.

Earnings / M&A / guidance: companySymbol with date or addDate sort; Stories when you need one cluster with top companies attached.

A grounded company news summary of these hits belongs on the Search Summarizer path - keep this guide on entity search, then summarize the filtered set.

When to use Articles vs Stories for company search

Need	Use	Entity scope
Every linked mention	Articles API (/v1/all)	Company or person on that article
One narrative per event	Stories API (/v1/stories/all)	Top companies / top people on the cluster
Headline-only surfaces	Headlines	Same company/person params as Articles

Clustering mechanics live in the Stories guide. Here the decision is only top-entity vs any linked mention.

Best practices: compose entities with other filters

Keep this guide focused on who. Around the entity allowlist, add the usual news search controls - source / sourceGroup, from / to / addDateFrom, topic (to filter news by topic), category, excludeLabel. A common business news provider pattern is companyDomain (or companySymbol for portfolios) + a trusted sourceGroup + addDateFrom for incremental pulls.

Entity filters ship on plans that include article and story search; compare tiers on API pricing.

Key takeaways

A company news API is entity-aware news search - companies and people as linked objects, not only keywords.

Use companyDomain, companySymbol, companyId, and companyName for orgs; personName and personWikidataId for executive mentions.

Articles return every linked mention; Stories scope to top companies and top people on the cluster.

Keep keywords for event language inside an entity allowlist.

Resolve IDs on /companies and /people, then reuse them in competitor, portfolio, and executive queries.

Get started: Sandbox · Entities · API reference · API pricing

Healthcare News: Filter by Perigon Trade Publication sources

Oleksyj Mokrousov — Fri, 24 Jul 2026 11:52:28 +0000

Healthcare news from trade publications — one filter, 35+ sources

Healthcare news spans hospital operations, payer strategy, pharma and biotech news, health IT, policy news, and medical devices. A keyword search alone pulls lifestyle blogs, press releases, and unrelated mentions. A filter news by source approach scopes results to trusted news sources you actually want in a filtered news feed.

Perigon's healthcareTradePublications source group is a ready-made publisher list for that job: 36 domains covering trade publication news across the full vertical. Add one parameter to the news API — sourceGroup=healthcareTradePublications — and every Articles, Headlines, or Stories search runs only against those outlets.

Try it live: Test healthcare news with sourceGroup in the sandbox · Read the docs: Sources and source groups · Compare plans: API pricing and features

What is the healthcareTradePublications source group?

Display name: Healthcare Trade Publications

Description: Healthcare-focused sources covering pharma, biotech, policy, hospitals, and medical devices.

Instead of hand-picking domains on every query, sourceGroup turns a curated bundle into a reusable filter — the same pattern as sourceGroup=top100, but scoped to vertical news and industry news sources for healthcare.

Core U.S. trade publications (12 domains)

Domain	Publication	Focus
modernhealthcare.com	Modern Healthcare	Hospitals, policy, finance, executives
beckershospitalreview.com	Becker's Hospital Review	Hospital operations, leadership
fiercehealthcare.com	Fierce Healthcare	Business, insurers, digital health
healthcaredive.com	Healthcare Dive	Policy, providers, industry trends
healthcareitnews.com	Healthcare IT News	EHR, AI, cybersecurity, interoperability
statnews.com	STAT	Biotech, pharma, medical research
medpagetoday.com	MedPage Today	Clinical research, physician news
managedhealthcareexecutive.com	Managed Healthcare Executive	Managed care, PBMs, payers
medcitynews.com	MedCity News	Health tech and startups
healthaffairs.org	Health Affairs	Policy research
healthleadersmedia.com	Health Leaders Media	Executive strategy, finance
healthtechmagazine.net	HealthTech Magazine	Hospital IT infrastructure

Pharma, biotech, and life sciences (10 domains)

Domain	Focus
genengnews.com	Genetic Engineering & Biotechnology News
european-biotechnology.com	European biotech industry
biotechniques.com	Life science methods and research tools
insideprecisionmedicine.com	Precision medicine
crisprmedicinenews.com	CRISPR and gene editing
bionity.com	Pharma and biotech (international)
pharmaphorum.com	Pharma industry commentary
contractpharma.com	Contract manufacturing and CDMO
analytica-world.com	Analytical science and lab industry
press.asimov.com	Biotech and synthetic biology

Digital health, policy, international, and adjacent sources (14 domains)

Domain	Focus
digitalhealthnews.com	Digital health industry
healthday.com	Clinical and consumer health syndication
who.int	World Health Organization — global policy
ncpa.org	National Community Pharmacists Association
bwhealthcareworld.com	India healthcare business
health.economictimes.indiatimes.com	India healthcare business (Economic Times)
pharmaceuticals.einnews.com	Pharma news aggregation
beckersdental.com	Dental practice and DSO industry
healthimpactnews.com	Advocacy health news
parentherald.com	Parenting and family health
med.stanford.edu	Stanford Medicine newsroom
uofmhealth.org	University of Michigan Health
newsroom.uw.edu	University of Washington medicine news
medangel.org	Medical philanthropy and health nonprofit news

Together these domains cover the lanes most teams need for healthcare news monitoring — not just the eight names AI assistants typically suggest.

Add sourceGroup to your news API search

The sourceGroup parameter works on the Articles API (/v1/all), Headlines API (/v1/headlines), and Stories API (/v1/stories/all). Pass the group name exactly as registered:

GET https://api.perigon.io/v1/all?apiKey=YOUR_API_KEY&sourceGroup=healthcareTradePublications&sortBy=addDate&size=25

That single line replaces a long source= whitelist and keeps your healthcare news feed consistent across apps, alerts, and dashboards.

For headline-only products:

GET https://api.perigon.io/v1/headlines?apiKey=YOUR_API_KEY&sourceGroup=healthcareTradePublications&sortBy=addDate

For story tracking instead of article floods — useful when FDA decisions or hospital mergers reprint across dozens of trade pubs:

GET https://api.perigon.io/v1/stories/all?apiKey=YOUR_API_KEY&sourceGroup=healthcareTradePublications&topic=Health&sortBy=addDate

The Stories API returns clustered narratives with story velocity, summaries, and related articles — so breaking news monitoring fires on storylines, not every wire reprint. See the Stories API guide for clusterId and expandArticles.

Why a source group beats a static "top 8" list

Ask GPT, Claude, or Gemini for top healthcare news outlets and each returns eight familiar names — but they disagree on half the list. All three agree on five core pubs (Modern Healthcare, Becker's Hospital Review, Fierce Healthcare, Healthcare Dive, Healthcare IT News). None surface the full pharma, biotech, international, or device trade depth in healthcareTradePublications.

Dimension	Typical AI "top 8"	healthcareTradePublications (36)
U.S. hospital/business core	Strong	Strong + extended
Payer / managed care	Sometimes	✓
Biotech / pharma depth	STAT at most	10+ dedicated sources
Health IT / digital	Partial	HITN, HealthTech, MedCity, digitalhealthnews
Policy / research	Sometimes	Health Affairs + WHO
International	Rarely	India, EU biotech
Medical devices / lab	Rarely	analytica-world, contractpharma

A source group solves that curation problem once. Your news API calls stay the same; you swap or extend groups without rewriting queries.

Healthcare news monitoring workflow

A typical media monitoring stack layers filters in this order:

Step	What it does	Parameter
1. Scope publishers	Lock to trade publications	`sourceGroup=healthcareTradePublications`
2. Match topics	Filter news by topic or keyword	`q`, `topic`, `companyDomain`
3. Time window	Filter news by date for digests	`from`, `to`, `addDateFrom`
4. Track narratives	Story tracking across reprints	Stories API + same `sourceGroup`
5. Brief leadership	News briefing from filtered corpus	Search Summarizer on the result set

Poll the Articles API with sourceGroup=healthcareTradePublications and addDateFrom set to your last checkpoint for real-time news alerts. Layer the Health taxonomy branch for news classification — see the news taxonomy guide.

When the feed is scoped but volume is still high, pass matches into Perigon's Search Summarizer for a grounded AI news summary. The briefing stays tethered to the news articles your sourceGroup filter selected.

Try it: Search Summarizer sandbox · Search Summarizer API reference

FAQ

What is healthcareTradePublications?

A Perigon source group containing 36 healthcare trade publication domains — hospitals, payers, pharma, biotech, policy, health IT, medical devices, and selected international outlets. Pass sourceGroup=healthcareTradePublications on the news API to scope searches to that bundle.

How do I filter healthcare news to trade publications only?

Add sourceGroup=healthcareTradePublications to any Articles, Headlines, or Stories API request. This is filter news by source at the publisher-bundle level — more reliable than keyword-only search for healthcare news from industry news sources.

Can I combine sourceGroup with keywords and dates?

Yes. sourceGroup sets the publisher boundary; add q, topic, companyDomain, from, to, and addDateFrom to narrow within the group. You can also use excludeSource to drop individual domains from an otherwise correct bundle.

How is this different from listing domains in source=?

sourceGroup is a maintained, reusable publisher list. You do not paste 36 domains into every query, and Perigon can update the group without changing your integration. For a custom subset, create a private source group and pass your group name the same way.

What are the best sources inside the group for healthcare news?

For U.S. business coverage: Modern Healthcare, Becker's Hospital Review, Fierce Healthcare, Healthcare Dive, and Healthcare IT News. For biotech and pharma: STAT and GEN. For policy: Health Affairs and WHO. The full group covers all lanes in one filter.

Historical News API Search Long-Range News Archives and Coverage Since 2016

Oleksyj Mokrousov — Thu, 23 Jul 2026 14:47:07 +0000

Historical news API - questions long-range archive search answers

Products marketed as media archive tools often stop at "filter the last 30 days." That fails the moment a team needs past news coverage across a merger, a regulation change, or a decade of company news history. Live feeds and bookmarked URLs miss old news articles that never appear in today's crawl. Compliance and research workflows need proof of what was published when - not just what was indexed yesterday. Teams also confuse publication time with ingestion time and pull the wrong window for retrospective news analysis.

Five questions come up on every historical news search project:

What is a historical news API in practice?
How can a news archive API search coverage going back years - including news since 2016?
How do publication date, add date, and refresh date differ?
How do you build timelines and run retrospective news analysis?
How do you export large news datasets for research or custom delivery?

Perigon answers these through the Articles API - one news API surface with archive-depth news data, not a separate historical product. Sections below follow that order.

Try it live: Test date filters in the Perigon sandbox · Read the docs: Dates and times · Compare plans: API pricing and features

What a historical news API means on the Articles API

A historical news API is the ability to search news articles across a deep back catalog - not a standalone endpoint. On Perigon, historical news lives in the same Articles API (/v1/all) used for breaking coverage: structured news article database records with headline, body, publisher, entities, topics, and multiple timestamps.

The historical news database is the indexed archive behind that endpoint. How far back you can query depends on your plan's historical lookback (see plans section below). Commercial tiers unlock 10+ years of historical news - aligning with news since 2016 for full-archive research, industry news archive work, litigation support, and longitudinal studies. Free and Basic tiers intentionally cap depth for exploration and early-stage products.

Every result is a full article object from the data model - suitable for dashboards, research tools, and news datasets pipelines without a second archive integration.

How to search news archive coverage by publication window

Long-range historical news search starts with when the story was published. On the Articles API, pass from and to to bound pubDate - the publication timestamp. Use ISO 8601 (date-only 2020-01-01 or datetime 2020-01-01T00:00:00) as described in Dates and times docs.

To study a policy debate across an election cycle, set from to the cycle start and to to election day, then add keywords or entity filters on the same request. To walk a timeline oldest-first, set sortBy to reverseDate. For newest-first archive browsing, use sortBy date.

Set showNumResults to true when you need the total match count before paginating - useful when sizing a media archive export or checking whether a news archive API query is too broad. Paginate with page and size when pulling large news datasets across months or years.

Your plan's historical lookback caps how far from can reach; requests beyond that window are rejected or clipped per account limits.

Publication date, add date, and refresh date

Three timestamps serve different jobs. Mixing them up is the most common bug in filter news by date workflows.

Timestamp	Params	Meaning	Typical use
Publication date	`from`, `to`	When the publisher dated the piece	Timelines, event history, past news coverage studies
Add date	`addDateFrom`, `addDateTo`	When Perigon first ingested the article	Catching newly discovered old news articles entering the index
Refresh date	`refreshDateFrom`, `refreshDateTo`	When Perigon last updated the record	Finding articles whose text or metadata changed after ingestion

Publication date answers "what was in the press in Q3 2019?" Add date answers "what entered our news data lake this week?" - including freshly backfilled archive pieces. Refresh date matters when summaries, entities, or body text were enriched after the first crawl; it usually tracks add date but can diverge for updated content.

Example of params (image 1):

Output example (image 2):

Sort options mirror those fields: date and reverseDate on publication time; addDate and reverseAddDate on ingestion; refreshDate on last system update. For live news search API polling, addDateFrom on your last checkpoint is the usual pattern. For historical news search, from/to on publication date is the default.

Timelines, retrospective analysis, and company news history

Retrospective news analysis combines a publication window with the same filters you use on live news - keywords (query logic), companyDomain, companyName, topics, sources, and taxonomy. That is how teams reconstruct company news history, compare executive mentions before and after an acquisition, or map industry news archive narratives across regulators and trade press.

Practical patterns:

Event history - from/to around the event, sortBy reverseDate, entity filter on the company or person involved.
Pre/post study - two publication windows with the same keyword set, compare showNumResults or exported pages.
Longitudinal desk - monthly from/to slices paginated into a warehouse for news datasets and trend charts.
Compliance review - publication window plus trusted source domains and label exclusions for press releases vs opinion.

The Stories API can extend the same date logic to clustered narratives on supported plans - useful when the unit of analysis is a developing storyline rather than every reprint.

News datasets, custom plans, and article dumps

Some projects outgrow interactive search. Research firms, risk teams, and data products need bulk historical news database extracts - millions of rows across thousands of entities.

Perigon supports custom plans and licensed article dumps tailored to scope, format, and rights. One engagement pulled articles data for 20,000 companies from 2016 onward - more than 60 million articles - for offline modeling and entity linkage. Smaller teams stay on standard API tiers and paginate the Articles API; large or recurring exports typically move to Commercial or Data License terms with dedicated delivery.

Talk to Perigon sales when you need custom lookback beyond standard tiers, entity-scoped bulk history, or redistribution rights for news datasets built on Perigon news data.

Historical depth by API plan

Archive reach scales with tier. Figures below match current public plan limits; confirm live values on the pricing page.

Plan	Historical lookback	Data freshness
Free	3 months	Real-time
Basic	6 months	1 hour delay
Plus	3 years	Real-time
Commercial	10+ years	Real-time

Commercial also adds Stories, Companies, People, Sources, and Journalists endpoints, custom sources, SLA, and dedicated support - the tier most historical news API products choose for news since 2016-scale research. Plus suits real-time products that still need several years of archived news articles. Free and Basic are for sandbox historical news search experiments within shorter windows.

Key takeaways: historical news search and news archives

Historical news API capability is archive-depth search on the Articles API - not a separate news archive API product.
Use from/to for past news coverage by publication date; addDateFrom/addDateTo and refresh params for ingestion and update windows.
sortBy reverseDate builds timelines; showNumResults sizes large news datasets before export.
Plan tier sets how far back historical news queries reach - up to 10+ years on Commercial, covering news since 2016.
Custom dumps and licensed extracts support enterprise retrospective news analysis at tens of millions of articles.

Get started: Try date filters in the sandbox · Dates and times guide · API reference · API pricing

Google Has No News API - Perigon Does

Oleksyj Mokrousov — Thu, 23 Jul 2026 14:35:20 +0000

What is a news API?

News monitoring usually means the same manual work every day: open multiple news sites, run the same searches, check whether anything new appeared, and copy links into a spreadsheet or Slack. That does not scale—articles get missed, filters change every time, and hours go into searching instead of acting on what you find.

A news API fixes that. You define what you care about once—keywords, publishers, date range, topics, companies—and your application fetches matching news articles automatically. Each result comes back as structured news data: title, URL, publication date, full text, publisher, topics, and mentioned people or companies.

A news search API (also called an article search API or news data API) is how you run those searches. One HTTP request replaces a morning of manual checking:

GET https://api.perigon.io/v1/all?apiKey=YOUR_API_KEY&q=Tesla&source=reuters.com&from=2026-07-01

You get every Reuters article about Tesla since July 1—ready for a filtered news feed, dashboard, or news alert. No repeated searches across sites.

Manual news monitoring	With a news API
Check dozens of sites by hand	One query across many publishers
Repeat the same search daily	Saved query runs on a schedule
Miss articles outside your routine	`addDateFrom` catches everything new since last check
Hard to filter by source, date, or company	`source`, `from`/`to`, `companyDomain` in the same call
Copy-paste into tools	Structured JSON goes straight into your app

Who uses a news API?

Teams reach for a news provider or news data provider when keywords matter and volume is too high to track by hand:

PR and comms — monitor brand mentions, executive names, and crisis keywords across publishers
Finance and investing — track ticker symbols, earnings, and M&A terms from trusted financial sources
Competitive intelligence — watch competitor domains, product launches, and industry topics daily
Research and legal — search news articles by body text over a date range, excluding opinion pieces
Product teams — embed a news feed provider or news aggregator inside an app or customer portal

In every case the pattern is the same: pick your keywords, lock in your sources, and let the API news layer do the searching.

How is this different from Google News or RSS?

Many developers search for a google news api. Google News is a website for reading headlines. It does not give you a documented API to search news articles by publisher, date range, and topic in one request.

Approach	What you get	Limitation
Google News (website)	Headlines for humans	No server API for filtered article search
RSS feeds	Updates from one publisher	One source at a time, limited filtering
Web search	Web pages of all types	News mixed with blogs, forums, old pages
Perigon News API	Structured articles + filters	Requires an API key; built for developers, with free test plans

RSS works if you follow five publishers. It breaks when you need news filtering across hundreds of sources by keyword, company, or date.

Web search returns everything - not just news articles - and you cannot reliably filter by reuters.com + from=2026-07-01 + companyDomain=tesla.com in one call.

Perigon is a news provider and news content provider: it indexes articles from many publishers, enriches them with metadata (topics, companies, locations), and exposes them through API news endpoints your backend calls directly. Think of it as a news article database you query with keywords instead of browsing.

Perigon's article endpoint: `/v1/all`

The main endpoint for searching individual articles is GET /v1/all (also available at /v1/articles/all). Pass search terms and filters as URL query parameters.

Base URL: https://api.perigon.io/v1

First request: try it in the Sandbox

GET https://api.perigon.io/v1/all?apiKey=YOUR_API_KEY&q=renewable+energy&size=10&page=0

Pass your API key as apiKey in the query string, or via the x-api-key header or Authorization: Bearer header.

What you get back

Every response has three parts: status (200 on success), numResults (total matches—enable with showNumResults=true), and articles (the list of results).

Each article includes the fields you need for feeds, and dashboards:

The article itself — title, description, content, url, imageUrl, and articleId
Who published it — source.domain, source.paywall, authorsByline, language, country
When — pubDate (published), addDate (indexed by Perigon)
What it's about — categories, topics, keywords, labels (e.g. Opinion, Press Release)
Who and what is mentioned — people, companies, locations
Extra signals — sentiment, clusterId (links related coverage), summary

That metadata is what turns a keyword search into something actionable. A PR team does not just need the word "recall" in a headline—they need to know the publisher, whether it is opinion, and which companies are mentioned. A news data API returns all of that in one object.

Full field list and live testing: API Reference

How do you use a news search API to find relevant articles?

News search is not just typing a keyword. The useful part is combining text queries (what the article says) with filters (who published it, when, about what). That is news filtering—and it is what separates a news search API from a basic headline list.

Text search: `q`, `title`, `desc`, `content`, `summary`, `url`

Parameter	What it searches	When to use it
`q`	Title + description + content	General keyword search (start here)
`title`	Headline only	Brand or executive mention monitoring
`desc`	Short description / lede	Matching article summaries
`content`	Full article body	Deep research queries
`summary`	Summary field	When summaries are available
`url`	Article URL path	Site sections, e.g. `source=cnn.com&url=travel`

All text fields support Boolean operators (AND, OR, NOT), exact phrases in quotes, and wildcards (*, ?).

# General search
GET /v1/all?q="supply chain" AND shipping&language=en

# Headlines only
GET /v1/all?title=Tesla&from=2026-07-01&sortBy=date

# Body text, excluding a term
GET /v1/all?content="generative AI" NOT OpenAI&category=Tech

Tip: use q for broad news search. Add title, desc, content, summary, or url when you need to search one part of the article only.

Real-world keyword patterns:

Goal	Query approach
Brand monitoring	`q="Acme Corp" AND (lawsuit OR recall OR CEO)`
Executive tracking	`title="Jane Doe"` — catches headlines only, less noise
Industry research	`content="data center" AND energy` + `category=Tech`
Site-section coverage	`source=cnn.com&url=travel&q=airline`

Date filters

Control when articles were published or added to Perigon:

Parameter	Meaning
`from` / `to`	Article publication date range
`addDateFrom` / `addDateTo`	When Perigon first indexed the article
`refreshDateFrom` / `refreshDateTo`	When Perigon last updated the article

Dates accept ISO (2026-07-01T00:00:00) or yyyy-mm-dd.

GET /v1/all?q=acquisition&from=2026-06-01&to=2026-07-06&sortBy=date

Use from/to for "articles published this week"
Use addDateFrom for "articles added since my last poll" (best for news alerts)

A weekly research report uses from/to. A Slack alert that fires every 15 minutes uses addDateFrom.

Source filters

Control where articles come from:

Parameter	Meaning
`source`	Publisher domain (e.g. `reuters.com`; wildcards like `*.cnn.com` supported)
`sourceGroup`	Curated publisher bundle (e.g. `top100`)
`excludeSource`	Remove specific publishers
`excludeSourceGroup`	Remove a publisher bundle

GET /v1/all?q=earnings&source=wsj.com&excludeSource=example-blog.com

This is core news filtering: combine keywords with trusted publishers instead of searching the entire web. A news feed provider that only shows top-tier sources is just sourceGroup=top100 on every call.

You can also filter by topic, category, companyDomain, personName, country, language, and label (e.g. exclude opinion or press release).

Pagination

Parameter	Default	Meaning
`page`	`0`	Which page to return (starts at 0)
`size`	`10`	How many articles per page
`showNumResults`	`false`	Set `true` to get total match count (up to 10,000)

GET /v1/all?q=electric+vehicles&page=0&size=25&showNumResults=true

Increment page to walk through a large result set—useful when building news datasets or export tools.

Sorting

`sortBy`	Order
`relevance`	Best match to your query (default)
`date` / `pubDate`	Newest published first
`reverseDate`	Oldest published first
`addDate`	Newest added to Perigon first
`reverseAddDate`	Oldest added first
`refreshDate`	Most recently updated first

Search box in your app → sortBy=relevance
Live filtered news feed or news alerts → sortBy=date or sortBy=addDate

How can an article search API support feeds, dashboards, or research tools?

The same /v1/all call with different keywords and filters powers very different products.

Filtered news feeds

A news feed provider ships a stream of articles to users based on saved interests:

Save a query — e.g. q=AI regulation&sourceGroup=top100&language=en
Call /v1/all every few minutes with sortBy=date
Show title, url, description, pubDate, and source.domain

Set showReprints=false to hide syndicated wire copies. Use clusterId to group articles about the same story so your feed does not show the same event ten times.

Example: a fintech app embeds a "regulatory news" tab. One saved query, one API call, always fresh news articles—no manual curation.

Dashboards

Different teams filter the same news data API differently:

Team	Filters	Why
Compliance	`excludeLabel=opinion` + `label=press release`	Official announcements only
Finance	`companySymbol=AAPL` + `topic=Markets`	Ticker-specific market news
Local news	`city=Chicago` or `state=IL`	Geography-based coverage
Search UI	`showHighlighting=true`	Bolded keyword snippets in results

One news aggregator backend, many dashboard views—each is a different keyword + filter combination.

Internal research

Researchers need depth, not speed. They use content or q with date windows (from/to), exclude opinion with excludeLabel=opinion, and paginate with page and size to review full result sets.

Example: a policy team studying climate legislation runs content="carbon tax" AND congress with from=2025-01-01, exports pages of results, and builds a news dataset for analysis—something no RSS reader or Google News tab can do with the same precision.

What should developers look for in a news provider or news data provider?

If you are evaluating a news provider, news data provider, or news content provider, check that the API offers:

Text search across fields — q, title, desc, content, summary, url
Rich filters — source, date, topic, category, entity, geography, language
Structured JSON responses — article ID, source, dates, entities, labels
Freshness timestamps — publication date plus ingestion date for polling
Pagination and sorting — page, size, sortBy, showNumResults

A good news feed provider or news aggregator backend needs all of these—not just a headline list.

Perigon vs the google news

Teams often start by Googling "google news api." Here is the short answer: Google does not offer a public news API. Google News is a website and mobile app for reading headlines—not a service your backend can call.

That does not mean Google News is useless. It is excellent at what it was built for. The gap appears when you need to automate news search inside your own product.

What you want to do	Google News (website)	Perigon `/v1/all` (API)
Browse today's headlines as a reader	Yes	Yes
Search by keyword in a browser	manually	programmatically
Limit results to specific publishers	UI only	Yes — `source`, `sourceGroup`, `excludeSource`
Filter by date range	UI only	Yes — `from`/`to`, `addDateFrom`
Poll for new articles on a schedule	No — you would have to check by hand	Yes — `addDateFrom` on each poll
Get full article text, companies, people, topics	No — headlines and snippets only	Yes — `content`, `companies`, `people`, `topics`
Build a feed or dashboard in your product	No — no way to connect it	Yes

Google News can help a person stay informed. Google News cannot send your server a filtered list of Reuters articles about Tesla published since July 1, updated every 15 minutes, with full text and entity metadata.

Perigon can. Your backend sends one request:

GET /v1/all?apiKey=YOUR_API_KEY&q=Tesla&source=reuters.com&from=2026-07-01&sortBy=date

You get JSON back. Paginate with page and size. Poll for new matches with addDateFrom. Pipe the results into a filtered news feed, a Slack news alert, or a research dashboard.

You are not replacing Google News for people who want to read the news. You are adding a news data API for products that need to search news articles automatically.

Local News API: Search Local News by City, State, County, Radius, and Source Location

Oleksyj Mokrousov — Wed, 22 Jul 2026 11:28:55 +0000

Local news API: the issues geography filters resolve

Location based news breaks down the moment a product treats place names as plain keywords. Portland matches cement trade coverage and travel blogs, not just Portland, Oregon. A headline that mentions Chicago is not the same as a story about Chicago. County papers and neighborhood sites rarely win broad text search, yet they are exactly what hyperlocal news readers want. National crawls cannot tell whether a Denver dateline means the event was in Denver or the newsroom is in Denver.

Those gaps show up as five practical problems teams hit when building on a news search API:

How do you scope city news search to the right market?
How do you filter regional news data by state or county?
How do you find hyperlocal news from local publishers?
How do you build local news near me and local news today feeds?
How do you separate where a story happened from source location?

Perigon's geography filters on the Articles API, Stories API, and Sources API address each problem in turn below.

Try it live: Test local news search in the Perigon sandbox · Read the docs: Local news guide · Compare plans: API pricing and features

City news search: scoping the right market

Place names repeat across countries and states, so keyword news search alone cannot tell which Portland or Springfield you mean. On the Articles API (/v1/all), pass city together with state to anchor urban city news to one market. When you need both values to match - not either/or - use structured location:

location=city:New York::state:NY

The Stories API (/v1/stories/all) accepts the same city and state values when you track storylines instead of individual articles.

Regional news data by state, county, and area

Statewide desks, county government coverage, and borough-level products need geography wider or narrower than a single city. Event-location params on the Articles API match articles where the named place is central to the story - not a passing mention.

Pass state for statewide regional news data, county when metro boundaries do not fit, and area for neighborhoods or districts. Multiple values on one param use OR logic. Exclude params such as excludeState drop specific markets from a national crawl without rebuilding the query. See location and geo filters for full event-location behavior.

Hyperlocal news and local publishers

Small outlets rarely surface in keyword results, but they are registered in Perigon's publisher index with city, state, county, country, and coordinates. Source location params on the Articles API pull articles from newsrooms in a market: sourceCity, sourceState, and source radius (sourceLat, sourceLon, sourceMaxDistance).

When the workflow starts with outlet selection, the Sources API (/v1/sources/all) returns local publishers as records - domain, name, geography - using the same source-location vocabulary. Whitelist those domains, then pull articles through the Articles API. See Sources and source groups for publisher discovery and curated bundles.

Local news near me and local news today feeds

Local news near me needs coordinates, not just city strings. Pass lat, lon, and maxDistance on the Articles API (kilometers, up to 300) to scope geospatial news around a point. Add a topic or keyword for vertical lanes - weather news along a storm track, for example.

Local news today feeds poll for fresh matches: addDateFrom on your last checkpoint, sortBy on addDate, and city or state for the market. Set type to local when Perigon's local classification should sit on top of geography. For storyline views instead of article floods, the Stories API clusters coverage by city, state, and area - summaries, velocity, and related articles in one object.

What you need	Endpoint	Geography params
Articles in a market	Articles API `/v1/all`	Event and/or source location, radius, `type`
Storylines in a market	Stories API `/v1/stories/all`	`city`, `state`, `area`
Headlines with geo scope	`/v1/headlines`	Same as Articles API
Publisher directory	Sources API `/v1/sources/all`	Source city, state, country, radius

Event location vs source location

The fifth problem is conceptual: where the story happened and where the publisher sits are different filters.

Axis	Question it answers	Example use case
Event location	Where did the story happen?	City council, weather emergencies, crime, school board
Source location	Where is the publisher based?	Hyperlocal reader apps, publisher directories, DMA ad products

A Miami Herald story about Tallahassee legislation is Tallahassee by event, Miami by source. Local news monitoring dashboards usually lead with event params; hyperlocal news apps often weight source params. Combine both on the Articles API - county set to Cook County with sourceState set to IL - when you need events in one county from Illinois-registered publishers.

Geospatial news: local news data on every article

Scoped results arrive with local news data your UI can render without extra geocoding: tagged places, publisher source location, topics, entities, sentiment, summaries, and story cluster linkage.

Local news monitoring: weather news and politics news patterns

Weather news and emergency monitoring

Layer state or county with storm-related keywords or a weather topic; add maxDistance when the threat moves across a map.

Regional politics news desk

Pass state, category Politics, and opinion label exclusions. Add sourceState when the desk tracks who published the piece, not only where the vote happened.

Multi-market desks

Run one scoped query per DMA - different city or sourceCity each - and aggregate into a regional news data store. Use showNumResults to spot volume spikes without full page downloads.

News search API: when to use event vs source filters

Goal	Start with	Add if needed
Right city, right state	`city` + `state` or structured `location`	Keywords, topics, dates
Statewide or county desk	`state` or `county`	Topic, category, labels
Outlets in a market	`sourceCity` on Sources API	Domain or group filters
Feed near a user	`lat`, `lon`, `maxDistance`	`q` or entities
Event vs newsroom geography	event params vs source params	Combine on Articles API

Stack geography filters on the news search API

Geography composes with keywords, source domains, date windows, topics, entities, and label rules on the Articles API. A crime desk might pass county, a Crime topic, a one-week date window, language, and showReprints false - filter news by location and subject in one request.

News API plans for local media monitoring

Geography filters on the Articles, Headlines, Stories, and Sources APIs are included across Perigon plans; volume and history depth grow with tier.

Plan	Best for	Highlights for local workflows
Free	Exploring the API	150 requests/month for sandbox geo experiments
Basic	Eligible startups	10,000 requests/month, 6 months of history
Plus	Higher-volume products	50,000 requests/month, real-time data, 3 years of history
Commercial	Production	100,000+ requests/month, 10+ years of history
Data License	Full commercial data products	Full license for apps built on Perigon news data

Compare plan features on the pricing page linked in the intro.

Key takeaways: local news near me and city news search

Geography filters on the Articles API, Stories API, and Sources API let you scope local news by city, region, publisher, and radius.
Structured location resolves ambiguous city news search; radius params power geospatial news and local news near me.
The Sources API discovers local publishers; Articles and Stories APIs filter coverage and narratives in those markets.
Event params answer where the story happened; source location params answer where the newsroom sits.
Stack geography with topics and dates for weather news and politics news workflows.

Get started: Try local news in the sandbox · Local news guide · API reference · API pricing

Articles API: Filter and Search News by Google Content Categories

Oleksyj Mokrousov — Wed, 22 Jul 2026 11:25:25 +0000

Articles API taxonomy filters beat noisy keyword search

Keyword news search is the right starting point when you know the exact phrase you want. It is the wrong tool when you need a whole industry desk - every banking story, every biotech development note, every sports trade rumor - without writing a hundred OR clauses and still catching opinion blogs, product reviews, and unrelated mentions.

That is where news taxonomy filtering comes in. Instead of guessing which strings appear in headlines, you filter news data by content categories: hierarchical paths that describe what an article is about. Perigon's Articles API applies Google Content Categories (Version 2) to every article - and supports the same paths on story search - so developers can build filtered news feeds, industry news monitors, and vertical news products with news filtering that stays on-category.

Try it live: Test article search in the Perigon sandbox · Read the docs: Articles API reference · Compare plans: API pricing and features

The theory: when taxonomy wins over keywords

News filtering with keywords alone treats every match as equal. Search for "bank" and you get central banking, river banks, bank robbery, and a fintech app's "data bank" feature. Search for "AI" and you pull product launches, research papers, policy hearings, and stock tips - all mixed together.

Taxonomy is built for structured industry or category scoping when text search is too noisy. You declare the kind of content you want - finance, health conditions, team sports, rock music - and the Articles API returns articles whose article categorization aligns with that path. Combine taxonomy filters with filter news by source, filter news by date, and Perigon's own news categories and news topics when you need both precision and control.

The result is cleaner news datasets, fewer false positives in news monitoring, and industry news filtering that scales without maintaining giant keyword blocklists.

How taxonomy filters work on the Articles API

Perigon does not offer a separate taxonomy product. Google content category filtering is built into the Articles API - the same article search API and news search API surface you use for keywords, publishers, dates, companies, and topics.

Every indexed article already carries hierarchical content categories from Google: full paths with confidence scores that describe industry, subject family, and content type - not just what keywords appear in the text. When you search articles, you filter on those paths. When you search stories, you can filter clustered narratives the same way.

That gives you three capabilities in one news data workflow:

Pre-classified articles - every article in the index carries Google taxonomy paths with confidence scores, so your product knows what kind of story it is before you display it.
Taxonomy filters on the Articles API - narrow article search to one or more exact Google category paths, or to a whole branch with prefix matching.
Taxonomy filters on the Stories API - apply the same category logic to clustered narratives when you track storylines, not individual reprints.

You are not maintaining a category index or running offline classifiers. You query news articles the way you already would - and taxonomy is a filter on top, with enriched category metadata on every result.

Taxonomy, category, and topic: three different filters

Perigon exposes three related but distinct dimensions. Using the right one keeps news classification precise.

Dimension	What it represents	Examples	Best for
Taxonomy	Google Content Category paths - industry and content-type hierarchy	Rock Music, Banking, Health Conditions, Team Sports	Industry news filtering, vertical desks, content-type scoping
Category	Perigon broad editorial themes derived from coverage	Tech, Politics, Sports, Business, Finance, Entertainment	Fast theme filters, dashboards, mixed-industry products
Topic	Granular Perigon topics (entity- and theme-level)	Markets, Cryptocurrency, College Sports, Crime	Filter news by topic when you need finer slices than category

Rule of thumb: use taxonomy when the Google path describes the industry or content family you want. Use category for quick Perigon theme buckets. Use topic for specific storyline types. Stack them when a product needs both - for example, all of finance plus a Markets topic, excluding opinion pieces.

Articles can carry multiple taxonomy paths. Google's V2 model returns every applicable path with reasonable confidence - for example both Science and Astronomy when both fit. A child category labeled Other under a parent indicates the content aligns with the broader parent more than any specific sibling. See the full category list for every exact string (all paths are English).

See what taxonomy filtering delivers

The clearest way to understand the value is to walk through what a customer actually gets back.

Imagine a music media product that wants rock music coverage - not every headline that mentions a guitar, and not celebrity gossip unless it is genuinely about music. In the sandbox, you filter English articles by the Google path for Rock Music under Arts and Entertainment, sort by publication date, hide reprints, and ask for a page of recent matches.

One result is a piece from a New Jersey radio station about Metallica closing their M72 world tour with a Black Sabbath tribute one year after Ozzy Osbourne's death. That is exactly the kind of story a rock vertical wants: tour news, tribute performance, venue and residency details - not a generic entertainment mention.

What Perigon returns on that article is the full news data package your product can surface without extra enrichment:

Editorial classification - Perigon category Entertainment, plus Google taxonomies for Rock Music (strongest match), Celebrities and Entertainment News, and Concerts and Music Festivals, each with a confidence score.
The story itself - headline, description, full article text, publication and indexing dates, URL, hero image, author byline, and language.
Publisher context - source domain, country, and localized publisher geography down to city and coordinates when available.
Entities and people - resolved names such as Metallica, Kirk Hammett, Robert Trujillo, and Ozzy Osbourne, with types and mention counts; linked people records where Wikidata IDs exist.
Keywords and summaries - weighted keywords (Black Sabbath, M72 Tour, Electric Funeral, London Stadium) plus short and long AI-generated summaries for cards and digests.
Sentiment - positive, negative, and neutral scores across the piece.
Story linkage - a cluster ID tying this article to related coverage, so your app can jump from one article to the broader narrative without rebuilding clustering yourself.

That is what article categorization buys you: the filter defines the vertical, and every match arrives ready for a feed, alert, or dashboard. Your customer sees rock coverage; your backend receives structured content categories, entities, and metadata in one response.

How Google content categories improve Articles API search results

Google news taxonomy paths are standardized, hierarchical, and maintained at scale - thousands of content categories covering finance, health, government, sports, entertainment, and more. That beats home-grown tag lists that drift out of date.

For the Articles API, the payoff is immediate:

Less keyword noise - a Funds category finds fund coverage without every headline that mentions "fund" in another sense.
Consistent article categorization across publishers - regional, trade, and wire news sources share the same path vocabulary.
Prefix discovery - filter by an entire Finance or Health branch without listing every leaf category underneath.
Composable filters - taxonomy works alongside keywords, publisher, date range, company, and topic in one filtered news feed query.

Perigon maps these paths onto indexed news articles and story clusters, so news classification is queryable at API speed - not something you run offline per article.

Filter articles by exact path or whole industry branch

On the Articles API, you have two taxonomy filter modes.

Exact path matching accepts one or more full Google category strings. Use this when you know the precise branch - banking versus mutual funds versus insurance - without pulling everything under Finance. A fintech desk might request only Banking and Investing Funds paths, limited to articles published since the start of the month, sorted newest first.

Prefix matching filters by the start of a path only. One prefix covers every category beneath it. A finance monitoring product might start with the entire Finance tree across top publishers in English, then layer a Markets topic or a company watchlist on top. Prefix matching is available on the Articles API; the Stories API accepts full taxonomy paths but not prefix - use exact paths on stories, or filter articles first and follow cluster links into narrative views.

What your product receives on each article

Beyond taxonomy, each article carries Perigon categories and topics, the publisher record, dates, full text, summaries, sentiment, people and companies, locations, keywords, labels, and cluster membership. Taxonomy scores help you rank, display category chips, or route articles into the right queue in your own topic classification pipeline.

Story-level taxonomy for narrative monitoring

For news monitoring at storyline scale, filter clustered stories by the same Google paths. Stories inherit content categories from their constituent articles.

A sports product might ask for team-sports stories covered by at least five unique publishers, updated recently, sorted by when the narrative last moved. Story results aggregate taxonomies across the cluster - useful when you care about momentum in a vertical, not every single reprint. Pair a Sports category with a finer Google sports path when you want Perigon's broad theme plus a specific subtree.

Story objects bring the same enrichment customers expect at article level - generated story names, summaries, key points, sentiment, entities, velocity signals, and the related articles inside the cluster.

How taxonomy improves news filtering compared with keywords alone

Approach	Strength	Weakness
Keywords only	Exact phrases, breaking names, rare terms	Homonyms, cross-industry noise, endless OR lists
Perigon category / topic	Fast, editorially familiar buckets	Less granular than full Google trees
Taxonomy	Industry-wide, hierarchical, standardized news categories	Requires knowing path strings (use Google's list)
Taxonomy + keywords	Precise vertical coverage with named entities	Slightly more complex queries

The best news filtering stacks layers: a taxonomy prefix for the vertical, keywords or company filters for the watchlist, trusted publishers, a publication or indexing date window, and label exclusions when you want reporting only.

A healthcare intelligence workflow might combine the full Health taxonomy branch with clinical-trial or FDA keywords, major publishers only, a six-month publication window, and no opinion label - yielding health-classified news articles without every lifestyle blog that casually mentions "health."

What is the best way to filter and search news by industry?

For most industry news and vertical news products, this pattern works well:

Start with a taxonomy prefix - one branch per industry lane: Finance, Health, Sports, Computers and Electronics, Law and Government.
Add Perigon category or topic when you need an editorial theme or storyline type on top of Google paths.
Layer source and date - filter news by source (specific domains or curated publisher bundles) and filter news by date (publication window for reports, indexing window for polling alerts).
Switch to exact paths when the prefix is too broad - banking only, not all of finance.
Choose stories when the product tracks narratives; choose articles when it needs every piece.

That is industry news filtering at API level: structured lanes, optional keywords, trusted news sources, fresh news data.

Examples by vertical

Try these patterns in the sandbox. Use exact path strings from the Google Content Categories V2 list.

Finance

Monitor banking and markets coverage across top publishers: Finance prefix, Markets topic, major source bundle, sorted by when articles were indexed. Drill into mutual-fund coverage only with the Investing Funds exact path, a July publication window, and press releases excluded.

Technology

Combine Google's Computers and Electronics branch with Perigon's Tech category and artificial-intelligence keywords in English for a technology news lane that stays in-sector.

Health

Healthcare news monitoring: Health taxonomy prefix, FDA or clinical-trial keywords, Entertainment category excluded, articles from the past six months - clinical and regulatory signal without wellness listicles.

Politics

Policy and government coverage: Law and Government prefix, Politics category, US country filter, sorted by publication date.

Sports

Team and league coverage as stories rather than article floods: team-sports taxonomy path, Sports category, minimum unique publisher count, recently updated narratives.

Business

Cross-industry business reporting: Business category plus Business and Industrial taxonomy prefix, limited to trusted financial publishers, newest first.

Entertainment and music (vertical feed)

Rock Music exact path, English language, reprints hidden, sorted by date - exactly the shape of query that returns tour and tribute coverage like the Metallica M72 finale story, with taxonomies, people, summaries, and cluster linkage included. Your customer gets a filtered news feed of rock coverage; your app gets news articles ready to render.

Industry monitoring (multi-vertical product)

Build separate filtered news feeds per customer vertical with the same Articles API:

Customer vertical	What to combine
Fintech desk	Finance prefix + Cryptocurrency topic + indexing window for alerts
Pharma intelligence	Health prefix + company domain + opinion excluded
Logistics SaaS	Business and Industrial prefix + supply-chain keywords
Sports media app	Sports stories + team-sports taxonomy path

Poll with an indexing-since timestamp for alerts; request total match counts and paginate when exporting news datasets.

API plans: start free, scale when you need more

Taxonomy filters on the Articles API and Stories API are part of Perigon's core news search API offering - the same products you use for keyword search, source filters, and date windows. Whether you are prototyping a filtered news feed or running industry news filtering in production, the right plan depends on volume, data freshness, history depth, and which endpoints you need.

Perigon offers Free, Basic, Plus, Commercial, and Data License tiers. Just getting started or operating at scale? Compare all features on the API pricing page - request limits, rate limits, max results per call, real-time vs delayed data, historical lookback, commercial license terms, and which API products are included on each plan.

Plan	Best for	Highlights relevant to taxonomy workflows
Free	Anyone exploring the API	150 requests/month - enough to test taxonomy, prefix, and category/topic combinations in the sandbox
Basic	Eligible startups	10,000 requests/month, 6 months of history, Articles and related endpoints
Plus	Eligible startups at higher volume	50,000 requests/month, real-time data, 3 years of history - suited to live news monitoring feeds
Commercial	All businesses	100,000+ requests/month, custom rate limits, 10+ years of history
Data License	Full commercial data use	Full commercial license for products built on Perigon news data

Across plans, the API surface includes Articles, Stories, Search Summarizer, Sources, Companies, People, Journalists, and Topics - add taxonomy filters on Articles, pull story clusters on Stories, discover category labels on Topics, and narrow publishers on Sources.

Start here: sign up on the Free tier, run taxonomy queries in the sandbox, then upgrade when your news filtering workload outgrows monthly limits or you need real-time news data and deeper archives.

Key takeaways

Perigon does not ship a standalone taxonomy API - Google content category filtering lives on the Articles API, with matching filters on the Stories API.
Use exact taxonomy paths for precise verticals, prefix matching for whole industry branches on article search, and taxonomy on stories for narrative-level news filtering.
Taxonomy, Perigon category, and topic solve different jobs - combine them with filter news by source and filter news by date for production filtered news feeds.
Indexed articles expose taxonomies with confidence scores on every result - use them for display, ranking, or routing inside your app.
Browse the complete path list in Google's Content Categories documentation (Version 2).

Get started: Try taxonomy filters in the sandbox · Read the Articles API reference · Compare API plans and permissions