If you've ever built a side project, you know the drill: you have an idea, you start building, and then you spend 2-3 hours just searching for the right API.
Is there a free weather API? Does it need an API key? What's the rate limit? Does it have a Python SDK?
I got tired of that workflow. So I built a dataset.
What I built
A curated CSV with 1,555 public APIs across 52 categories, enriched with the information you actually need before integrating an API:
| Field | Example |
|---|---|
| Name | OpenWeatherMap |
| Category | Weather |
| Description | Real-time weather data by city or coordinates |
| Base URL | api.openweathermap.org |
| Auth Type | API Key |
| Free Tier | Yes |
| Free Limit | 1,000 req/day |
| Available SDKs | Python, JavaScript, Ruby |
| Docs URL | openweathermap.org/api |
| Popularity | High |
Why a CSV and not a website?
Websites go down. Datasets don't.
You can open this in Excel, Google Sheets, Notion, or Airtable. Filter by category, sort by auth type, search by SDK. No login, no account, no dependencies.
What categories are covered?
The 15 largest categories:
- Development (154 APIs)
- Geocoding (94 APIs)
- Games & Comics (89 APIs)
- Government (83 APIs)
- Transportation (66 APIs)
- Cryptocurrency (55 APIs)
- Finance (55 APIs)
- Video (50 APIs)
- Documents & Productivity (48 APIs)
- Social (39 APIs)
- Security (37 APIs)
- Sports & Fitness (37 APIs)
- Weather (36 APIs)
- Machine Learning (30 APIs)
- Music (32 APIs)
Plus 37 more categories including Animals, Anime, Books, Calendar, Email, Environment, Health, Jobs, Photography, Shopping, and more.
Some APIs you'll find inside
A few examples across different categories:
Weather: Open-Meteo (unlimited, no auth), WeatherAPI, OpenWeatherMap
Finance: Alpha Vantage, Fixer, ExchangeRate API
Machine Learning: Hugging Face, Stability AI, OpenAI
Geocoding: Mapbox, OpenStreetMap, HERE
Government: NASA, data.gov, EU Open Data Portal
Games: RAWG, IGDB, Open Trivia Database
How I built it
I scraped the public-apis community repository (one of the most maintained API directories on GitHub with 300k+ stars), then enriched each entry using Claude AI to fill in the missing fields: free limits, SDK availability, and popularity scoring.
The whole pipeline took about 3 hours of scripting and $0.02 in API costs.
Who is this for?
- Developers prototyping apps who need a quick API reference
- Indie hackers validating ideas before committing to a paid API
- Students building portfolio projects with real data
- No-code builders connecting tools via Make, Zapier, or n8n
- Data scientists looking for free data sources
Get the dataset
The full dataset (1,555 APIs, CSV format) is available here:
👉 DataStackHQ on Payhip — $9 one-time, instant download
If you find it useful, I'll be releasing more datasets. Drop a comment with what category you'd find most useful next.
Built with Python, pandas, and the Anthropic API. The source repository is public-apis/public-apis on GitHub.
Top comments (0)