v19 Β· Agentic AI Β· Runs locally Β· No server required

The AI Agent Studio
That Actually Does Things

BroClaw executes multi-step agentic tasks right in your browser β€” search the web, open pages, fix code, save full projects β€” without leaving your machine.

No signup Β· No subscription Β· API keys never leave your device Β· Auto-saves chats

BroClaw Agent β€” Task Running live
β–Ά Research the latest AI frameworks and build a comparison dashboard
πŸ” Searching: "best AI frameworks 2025"…
🌐 Opening: huggingface.co/blog β€” reading content…
🌐 Opening: github.com/trending β€” scanning repos…
πŸ” Searching: "LangChain vs LlamaIndex vs AutoGen 2025"…
πŸ’‘ Synthesising β€” 3 frameworks compared, 8 sources read…
πŸ“ Writing: /workspace/ai-comparison/index.html
πŸ“ Writing: /workspace/ai-comparison/style.css
πŸ“ Writing: /workspace/ai-comparison/app.js
βœ… Task complete β€” Dashboard saved & ready to preview

Everything You Need to
Actually Get Things Done

πŸ€–

Agentic Task Loop

BroClaw runs until the job is done. No fixed step limits β€” the AI decides what to do next: search, read, write, test, iterate β€” until your task is complete.

🌐

Real Web Search

Connects to Brave Search, SerpAPI, or your own SearXNG instance. Opens any URL in a sandboxed reader and extracts clean content automatically for the agent to reason over.

πŸ“

Save Code Projects

Generates full multi-file projects β€” HTML, CSS, JS, Python, JSON β€” and saves them directly to your local disk via the File System Access API. Preview instantly in-app.

🧠

Extended Thinking

Claude's extended thinking and o3/o4-mini reasoning tokens are fully streamed and displayed in real time so you can follow the agent's logic step by step.

πŸ”’

100% Private

Runs entirely in your browser. API keys are AES-GCM encrypted in localStorage and never sent to BroClaw servers. Your chats and files stay on your own machine.

⚑

Multi-Provider AI

Switch between Anthropic, OpenAI, Google Gemini, NVIDIA NIM, or any Ollama/LM Studio endpoint β€” all from one interface, mid-task with no restart required.

🎯

Skills

Save custom Skills that teach the agent your domain, tone, and preferred output formats β€” reusable across any task.

πŸ–₯️

Live Code Preview

Preview generated websites and apps directly inside BroClaw. Edit, re-run the agent, and see changes instantly without switching between tools.

πŸ”„

Ollama & Local Servers

Connect to any local Ollama instance, LM Studio, or vLLM endpoint. Full model routing with automatic fallback β€” switch providers mid-task with no restart.

Models That Fit Every Budget

ModelSpeedVRAM / RAMBest for
SmolLM2 1.7B⚑⚑⚑ Ultra2 GBQuick answers
Qwen3 0.6B⚑⚑⚑ Ultra1 GBFast tasks
Qwen3 1.7B⚑⚑ Fast2 GBCoding
Gemma 3 1B⚑⚑ Fast1.5 GBGeneral use
Llama 3.2 3B⚑ Balanced2.5 GBInstructions
Phi-4 Mini⚑ Balanced3 GBReasoning
Gemma 3 4BπŸ”΄ Powerful4 GBQuality output
Qwen3 4BπŸ”΄ Powerful5 GBComplex tasks

Downloaded once and cached in your browser β€” no re-download needed.

πŸ”‘
Keys stay local AES-GCM encrypted in your browser
🚫
No BroClaw servers Direct connection to AI providers
πŸ’Ύ
Chat saved locally IndexedDB in your browser only
πŸ“Š
Anonymous analytics Visit counts only, opt-in

Ready to run your first agentic task?

No signup. No credit card. Open the app and start in 30 seconds.

🦞 Open BroClaw

Desktop Chrome 113+ or Edge 113+ required for WebGPU local models