TokenSaver
EN

La documentation est disponible en anglais.

Documentation / Connect tools

LangGraph

Same LLM client wiring as LangChain — graphs call TokenSaver-backed models.

Follow the steps below in order. When you are done, send a test prompt and confirm a row appears in Flux IA for your ts_… key.

Integration pattern
  1. LangGraph
  2. LLM node
  3. TokenSaver egress
  4. Graph run

Steps

  1. Configure the graph LLM with OpenAI-compatible base_url
  2. Use ts_… as api_key
  3. Run the graph
  4. Inspect tool/LLM hubs in the agentic graph

Need more control? See Build paths.