Installation¶
Get Neurosurfer installed and confirm it can reach a model.
Minimal walkthrough¶
Install the core, plus any extras you need (imports are lazy, so the base install stays light):
pip install -U neurosurfer
# common combo: web search + gateway
pip install -U "neurosurfer[search,serve]"
Point it at a model via the environment (here, a local OpenAI-compatible server):
export LLM_PROVIDER=openai
export OPENAI_BASE_URL=http://localhost:1234/v1
export MODEL=qwen/qwen3.5-9b
Verify the setup — doctor reports the resolved provider and whether the endpoint answers:
Full notebook¶
The Colab notebook walks through every extra and both cloud (Anthropic/OpenAI) and local setups.
Next: Installation reference · Configuration · Tutorial 1 →