browser-use with an uncensored model
To run browser-use with an uncensored model, give its ChatOpenAI class a base URL and an abliterated model id, then hand it a local Chrome.
import os
from browser_use import Agent, Browser, ChatOpenAI
agent = Agent(
"Open http://localhost:3000, my own app, and list every page that shows an error.",
llm=ChatOpenAI(
"kimi-k3.1-pro-uncensored",
base_url="https://api.silkcompute.com",
api_key=os.environ["SILK_API_KEY"],
),
browser=Browser("chromium"),
)
print(agent.run_sync().final_result())- OpenAI format
- ChatOpenAI(base_url=)
- Anthropic format
- Not documented
- Tool calling
- Custom tools
- Streaming
- Not documented
- Browser use
- Local or cloud Chrome
- Memory
- Not documented
Labels link to browser-use's own docs, read November 18, 2026. Compare all 11
Point the agent only at systems, files and accounts you own or are allowed to test.
Deploy it
3 steps, with Kimi K3.1 Pro.
- 01
Install an exact version. The README asks for an exact pin, because releases can break APIs.
Terminalpip install "browser-use==1.0.6" export SILK_API_KEY=... - 02
Use ChatOpenAI with a base URL. Browser("chromium") opens a fresh local Chrome with a temporary profile.
See run.py above - 03
Keep it on your own sites. Pass passwords as sensitive_data. The README warns it is not a sandbox.
Terminalpython run.py
Which model
Kimi K3.1 Pro
Long runs, image and video input
- Id
kimi-k3.1-pro-uncensored- Per 1M in / out
- $3.50 / $8.00
DeepSeek V4.1 Flash 1017
Value, image input
- Id
deepseek-v4.1-flash-1017-uncensored- Per 1M in / out
- $0.24 / $0.90
The config uses the Silk Compute base URL, which serves both models with a 1,048,576-token context.