Langfuse
Open-source platform for tracing, prompt management and evaluation of LLM apps, self-hosted or as a cloud service.
- Open source
- Can run on your own servers
- Developer API
About LangSmith
Tracing, evaluation and monitoring platform for agents and LLM apps, usable with or without LangChain.
LangSmith is a platform from the LangChain team for seeing and testing what LLM apps and agents do. Every run is recorded as a trace that shows each model call, tool call and intermediate step, with inputs, outputs, timing and token counts, so you can find where an agent went wrong.
Beyond tracing, it supports datasets of test cases, evaluations that score outputs with code or with a model acting as a judge, annotation queues where people review and label runs, and dashboards for monitoring live traffic. It works with LangChain and LangGraph but also with other frameworks and plain SDK calls.
Teams use it while building an agent to compare prompts and models, and after launch to watch quality and cost. Check where trace data is stored, how long it is kept and whether sensitive inputs need to be masked before they are logged.
Details
- Open sourceNo
Developer API
Yes
Not enough info yet on can run on your own servers, works with MCP. Check the tool's own site.
Listing written by WithAutobot, checked 2026-11-04. Report a correction or request removal. Back to agent observability and evaluation.