Views
No views yet
1from dotenv import load_dotenv
2from browser_use import Agent, ChatBrowserUse
3load_dotenv()
4
5llm = ChatBrowserUse(
6 model='browser-use/bu-30b-a3b-preview', # BU Open Source Model!!
7)
8
9agent = Agent(
10 task='Find the number of stars of browser-use and stagehand. Tell me which one has more stars :)',
11 llm=llm,
12 flash_mode=True
13)
14agent.run_sync()pip install vllm --upgrade1vllm serve browser-use/bu-30b-a3b-preview \
2 --max-model-len 65536 \
3 --host 0.0.0.0 \
4 --port 80001from dotenv import load_dotenv
2from browser_use import Agent, ChatOpenAI
3load_dotenv()
4
5llm = ChatOpenAI(
6 base_url='http://localhost:8000/v1',
7 model='browser-use/bu-30b-a3b-preview',
8 temperature=0.6,
9 top_p=0.95,
10 dont_force_structured_output=True, # speed up by disabling structured output
11)
12
13agent = Agent(
14 task='Find the number of stars of browser-use and stagehand. Tell me which one has more stars :)',
15 llm=llm,
16)
17agent.run_sync()| Property | Value |
|---|---|
| Base Model | Qwen/Qwen3-VL-30B-A3B-Instruct |
| Parameters | 30B total, 3B active (MoE) |
| Context Length | 65,536 tokens |
| Architecture | Vision-Language Model (Mixture of Experts) |