Open-source Python library for AI browser agents, plus Browser Use Cloud: stealth browsers and a hosted agent API.
Python ai-agentsai-toolsbrowser-agentbrowser-automationbrowser-usecdpcloud-browsercomputer-useheadless-browserllmmcpplaywrightpythonremote-browserstealth-browserweb-automationweb-scraping
docs: point bu-30b-a3b-preview at self-hosting instead of Cloud
MagMueller6f8b0626b6db
diff --git a/browser_use/llm/browser_use/chat.py b/browser_use/llm/browser_use/chat.pyindex c2f29da07..48b186a57 100644--- a/browser_use/llm/browser_use/chat.py+++ b/browser_use/llm/browser_use/chat.py@@ -62,7 +62,9 @@ class ChatBrowserUse(BaseChatModel):- 'bu-2-0-mini-preview': Cheaper and faster per token, opt-in while in preview- 'bu-1-0': Previous generation model, redirected to bu-2-0 at the gateway- 'bu-qa-1': Website QA model (tests a site and scores functionality/aesthetics)- - 'browser-use/bu-30b-a3b-preview': Browser Use Open Source Model+ - 'browser-use/bu-30b-a3b-preview': open weights, self-hosted only. Browser Use+ Cloud does not serve it; run it yourself with vLLM and point `ChatOpenAI` at+ your own endpoint. See examples/models/bu_oss.py.- Provider-prefixed ids resolved by the gateway, e.g. 'anthropic/claude-sonnet-4-6','openai/gpt-5.5', 'google/gemini-3-pro'.api_key: API key for browser-use cloud. Defaults to BROWSER_USE_API_KEY env var.diff --git a/examples/models/bu_oss.py b/examples/models/bu_oss.pyindex 9a6355e5f..803d61c38 100644--- a/examples/models/bu_oss.py+++ b/examples/models/bu_oss.py@@ -1,12 +1,23 @@-"""+"""Run the open-weights Browser Use model on your own GPU.++`browser-use/bu-30b-a3b-preview` is published under+https://huggingface.co/browser-use/bu-30b-a3b-preview. It is open weights you host+yourself, not a model Browser Use Cloud serves for you, so it needs no+BROWSER_USE_API_KEY and has no per-token price.+Setup:-1. Get your API key from https://cloud.browser-use.com/new-api-key-2. Set environment variable: export BROWSER_USE_API_KEY="your-key"+1. pip install vllm+2. vllm serve browser-use/bu-30b-a3b-preview --max-model-len 65536 --host 0.0.0.0 --port 8000+3. python examples/models/bu_oss.py++Point BU_OSS_BASE_URL at the server if it is not on localhost."""+import os+from dotenv import load_dotenv-from browser_use import Agent, ChatBrowserUse+from browser_use import Agent, ChatOpenAIload_dotenv()@@ -17,9 +28,11 @@ try:except ImportError:pass-# Point to local llm-use server for testing-llm = ChatBrowserUse(- model='browser-use/bu-30b-a3b-preview', # BU Open Source Model!!+# Any OpenAI-compatible server works; vLLM is what the model card recommends.+llm = ChatOpenAI(+ model='browser-use/bu-30b-a3b-preview',+ base_url=os.getenv('BU_OSS_BASE_URL', 'http://localhost:8000/v1'),+ api_key=os.getenv('BU_OSS_API_KEY', 'not-needed'),)agent = Agent(diff --git a/skills/open-source/references/models.md b/skills/open-source/references/models.mdindex 497d6fbbb..d8c63b395 100644--- a/skills/open-source/references/models.md+++ b/skills/open-source/references/models.md@@ -71,7 +71,28 @@ llm = ChatBrowserUse(model='bu-2-0-mini-preview') # Cheaper per token, opt-in| bu-2-0 (default, premium) | $0.60 | $0.06 | $3.50 || bu-2-0-mini-preview (opt-in) | $0.15 | $0.15 | $1.50 || bu-1-0 (redirects to bu-2-0) | $0.60 | $0.06 | $3.50 |-| browser-use/bu-30b-a3b-preview (OSS) | — | — | — |++### Open weights (self-hosted)++`bu-30b-a3b-preview` is our open-weights browser model: 30B total, 3B active, 64K context.+Browser Use Cloud does not route it, so there is no `bu-*` alias and no per-token price.+Run it yourself and talk to it with `ChatOpenAI`:++```bash+vllm serve browser-use/bu-30b-a3b-preview --max-model-len 65536 --host 0.0.0.0 --port 8000+```++```python+from browser_use import ChatOpenAI++llm = ChatOpenAI(+ model='browser-use/bu-30b-a3b-preview',+ base_url='http://localhost:8000/v1',+ api_key='not-needed',+)+```++Weights: https://huggingface.co/browser-use/bu-30b-a3b-preview## OpenAI