DEV Community: Rohit Raj

NVIDIA LocateAnything-3B: The Open Visual Grounding Model That Beats YOLO (2026 Guide)

Rohit Raj — Mon, 06 Jul 2026 05:54:29 +0000

NVIDIA quietly shipped LocateAnything-3B on May 26, 2026 — a 3B open-weights vision-language model that turns a plain-English phrase like "the submit button" into exact pixel boxes, no fixed class list, no retraining. It grounds objects, GUI elements, and text with up to 2.5x higher throughput than older box-by-box decoders. By early July it had crossed 1.2M Hugging Face downloads. Here is what actually changed, runnable code to try it, how it stacks up against YOLO / Grounding DINO / Florence-2 / Qwen2.5-VL, and the license catch that will stop you shipping it to production if you are not careful.

Read the full version with code samples, diagrams, and architecture details: NVIDIA LocateAnything-3B: The Open Visual Grounding Model That Beats YOLO (2026 Guide)

More engineering notes: rohitraj.tech/en/notes

Safari MCP Server: Apple\'s Official Debugger vs the Community Tools (2026 Guide)

Rohit Raj — Sat, 04 Jul 2026 06:40:18 +0000

Apple shipped an official Safari MCP server in Safari Technology Preview 247 on July 1, 2026 — 16 built-in tools that let AI coding agents see your rendered page, read the console, and evaluate JS directly instead of you describing screenshots. There\'s also a separate, older community safari-mcp npm ecosystem with 80+ tools that works on production Safari. Here\'s what\'s actually new, how to install either one, when Apple\'s version is the right pick, and how I\'d wire it into a real dev workflow without waiting for a CI runner that supports Safari headless.

Read the full version with code samples, diagrams, and architecture details: Safari MCP Server: Apple\'s Official Debugger vs the Community Tools (2026 Guide)

More engineering notes: rohitraj.tech/en/notes

Strix: The Open-Source AI Pentester That Proves Every Bug (2026 Guide)

Rohit Raj — Fri, 03 Jul 2026 06:18:47 +0000

Strix (usestrix/strix) hit #1 on GitHub Trending on July 3, 2026 with +2,137 stars in a day — 32.8k total, Apache 2.0. It runs autonomous AI agents that act like real hackers: they exploit your app, validate each finding with a working proof-of-concept, and file only bugs they actually broke. On the XBEN benchmark it solved 100/104 web challenges (96%) at ~$3.37 each. This is the builder\'s read — what it is, how to install and run it, whether it hallucinates, how it stacks up against XBOW and PentAGI, when to skip it, and exactly how I\'d wire it into a real MVP\'s CI pipeline without it torching your API budget.

Read the full version with code samples, diagrams, and architecture details: Strix: The Open-Source AI Pentester That Proves Every Bug (2026 Guide)

More engineering notes: rohitraj.tech/en/notes

Baidu Unlimited-OCR: The Open-Source Model That Reads 40+ Page Documents in One Pass (2026)

Rohit Raj — Thu, 02 Jul 2026 16:07:58 +0000

Baidu open-sourced Unlimited-OCR on June 22, 2026 (MIT) — a 3B mixture-of-experts model with 500M active params that parses 40+ page documents in a single forward pass. Its new Reference Sliding Window Attention (R-SWA) keeps the KV cache flat, so memory and latency stay constant as output grows. It scores 93.23 on OmniDocBench v1.5 — beating DeepSeek-OCR by 6.22 points — at 12.7% higher throughput. This is the builder\'s read: what R-SWA actually does, how to run it locally with Transformers and vLLM, where it beats a cloud OCR API, when to skip it, and exactly how I\'d wire it into a production RAG ingestion pipeline.

Read the full version with code samples, diagrams, and architecture details: Baidu Unlimited-OCR: The Open-Source Model That Reads 40+ Page Documents in One Pass (2026)

More engineering notes: rohitraj.tech/en/notes

Ornith-1.0: The Self-Improving Open-Source Coding Model, Tested (2026)

Rohit Raj — Wed, 01 Jul 2026 16:21:06 +0000

DeepReinforce shipped Ornith-1.0 on June 25, 2026 — an MIT-licensed family of coding models that learn to write their own agentic scaffold during RL instead of using a human-designed harness. The 397B flagship hits 82.4 on SWE-bench Verified (DeepReinforce reports it edges past Claude Opus 4.7); the 9B runs on a single 24GB card. This is the builder\'s read: what self-scaffolding actually is, real vLLM and Ollama run commands, an honest comparison table, when to skip it, and the chat-template gotcha that will send your local copy into a runaway loop.

Read the full version with code samples, diagrams, and architecture details: Ornith-1.0: The Self-Improving Open-Source Coding Model, Tested (2026)

More engineering notes: rohitraj.tech/en/notes

This Week in AI Dev: The Local-Agent Stack Went Production-Real (Week 27 of 2026)

Rohit Raj — Tue, 30 Jun 2026 05:58:25 +0000

Week 27 of 2026 in AI dev tools: DeepReinforce's Ornith-1.0 ships MIT-licensed self-scaffolding coding models from 9B to 397B, Qwen 3.6 27B becomes the local-dev sweet spot at 28GB, vLLM turns one API call into a bounded multi-model collaboration, Herdr multiplexes 15+ coding agents in your terminal, Wayfinder routes deterministically between local and hosted LLMs, and Anthropic ships Claude Tag for async Slack delegation.

Read the full version with code samples, diagrams, and architecture details: This Week in AI Dev: The Local-Agent Stack Went Production-Real (Week 27 of 2026)

More engineering notes: rohitraj.tech/en/notes

Best Open-Source Deep Research Agent to Self-Host in 2026 (Onyx vs DeerFlow vs Perplexica)

Rohit Raj — Mon, 29 Jun 2026 03:44:39 +0000

An open-source deep research agent now sits at #1 on DeepResearch Bench — ahead of OpenAI, Gemini, and Perplexity. So you no longer have to rent deep research from a frontier lab. This is the builder\'s read on the four worth self-hosting in 2026 — Onyx, DeerFlow 2.0, Perplexica/Vane, and Khoj — with live star counts, a runnable Docker self-host, an honest comparison table, when to skip self-hosting entirely, and the production wiring the READMEs leave out.

Read the full version with code samples, diagrams, and architecture details: Best Open-Source Deep Research Agent to Self-Host in 2026 (Onyx vs DeerFlow vs Perplexica)

More engineering notes: rohitraj.tech/en/notes

Microsoft Agent Framework vs LangGraph vs CrewAI: Which to Use Now That AutoGen Is Dead (2026)

Rohit Raj — Sun, 28 Jun 2026 03:25:38 +0000

AutoGen is in maintenance mode — Microsoft folded it and Semantic Kernel into the new Microsoft Agent Framework, which hit 1.0 GA in 2026. So the old "LangGraph vs CrewAI vs AutoGen" advice is stale. This is the builder\'s read: the same agent written in all three frameworks, where each one actually wins, an honest comparison table, how to migrate an AutoGen AssistantAgent to a ChatAgent, when to skip the Microsoft stack entirely, and the setup I\'d ship to production.

Read the full version with code samples, diagrams, and architecture details: Microsoft Agent Framework vs LangGraph vs CrewAI: Which to Use Now That AutoGen Is Dead (2026)

More engineering notes: rohitraj.tech/en/notes

Gemini Computer Use vs Claude vs OpenAI: Best Browser Agent 2026

Rohit Raj — Sat, 27 Jun 2026 18:49:26 +0000

Google baked computer use into Gemini 3.5 Flash on June 24, 2026 — a vision-based agent that clicks, types, and scrolls across browser, mobile, and desktop. This is the builder\'s read: what actually shipped, the real interactions.create agent-loop code, honest OSWorld numbers (Gemini 78.4 vs GPT-5.5 78.7 vs Claude Opus 4.8 83.4), a side-by-side against Claude computer use and OpenAI, when to skip it, and exactly how I\'d wire one into production without it draining a credit card or running a prompt injection.

Read the full version with code samples, diagrams, and architecture details: Gemini Computer Use vs Claude vs OpenAI: Best Browser Agent 2026

More engineering notes: rohitraj.tech/en/notes

GLM-5.2 vs Claude Opus 4.8: Should You Switch Your Coding Agent? (2026)

Rohit Raj — Fri, 26 Jun 2026 05:11:22 +0000

Z.ai shipped GLM-5.2 as open weights (MIT) in June 2026, and it matches Claude Opus 4.8 on real coding-agent tasks at a fraction of the per-token price. This is the builder\'s read: what actually shipped, the real code to call it and drop it into Claude Code, an honest cost breakdown (the per-token gap is huge but GLM burns ~3.3x more tokens), a side-by-side table, when to stay on Opus, and the hybrid routing setup I\'d actually ship — Opus for the 20% of tasks where the gap bites, GLM-5.2 for the other 80%.

Read the full version with code samples, diagrams, and architecture details: GLM-5.2 vs Claude Opus 4.8: Should You Switch Your Coding Agent? (2026)

More engineering notes: rohitraj.tech/en/notes

Gemini Interactions API: The Migration Guide from generateContent (2026)

Rohit Raj — Thu, 25 Jun 2026 01:56:49 +0000

Google made the Gemini Interactions API generally available in 2026 and quietly made it the default interface for Gemini models and agents. The core method is interactions.create, and it keeps conversation state server-side via previous_interaction_id instead of resending the full history every turn like generateContent. This is the builder\'s migration read: the actual code diff from generate_content to interactions.create, the interaction-scoped-tools footgun the docs bury, a side-by-side against generateContent and OpenAI\'s Responses API, an honest "when to stay on generateContent," and exactly how I\'d wire it into production with a fallback and a store=false privacy path.

Read the full version with code samples, diagrams, and architecture details: Gemini Interactions API: The Migration Guide from generateContent (2026)

More engineering notes: rohitraj.tech/en/notes

Mistral OCR 4 vs AWS Textract vs Google Document AI: The Cheapest Accurate Document API (2026)

Rohit Raj — Wed, 24 Jun 2026 03:49:02 +0000

Mistral shipped OCR 4 on June 23, 2026 — model mistral-ocr-latest — and it tops OlmOCRBench at 85.20, handles 170 languages, and costs $4 per 1,000 pages ($2 batch) against AWS Textract\'s $65 per 1,000 for forms-and-tables. Every comparison guide currently ranking still covers OCR 3 or ignores Mistral entirely. This is the builder\'s read: what actually changed in OCR 4, the API call with the new confidence-score gating, an honest accuracy-and-price table against Textract, Google Document AI, and Azure, where each one genuinely wins, when you should NOT pick Mistral, and exactly how I\'d wire it into a RAG ingestion pipeline in production.

Read the full version with code samples, diagrams, and architecture details: Mistral OCR 4 vs AWS Textract vs Google Document AI: The Cheapest Accurate Document API (2026)

More engineering notes: rohitraj.tech/en/notes