Langfuse
Public facts onlyOpen source core, paid extras. The core is under an OSI-approved licence; some features ship in separately licensed commercial add-ons. Not fully open source. Source code · Website
Records the requests your AI app sends to its models and the answers it gets back, so a team can debug problems, manage prompts and rate answer quality. (reviewed by BenchLytix)
Licence, summarised from its LICENSE file: MIT outside the ee/ directories, which are under a commercial licence. LICENSE file, read 2026-10-13.
Running it yourself
Its compose file names these supporting services: clickhouse, minio, redis. Services like these usually need more memory than a small server; check the project's own hosting requirements before you size one. This leaves out the time it takes you to run it. Compose file
What this label rests on
Licence, last update and source code, from public records. We have not run it.
Facts and where they come from
| Fact | Value | Source |
|---|---|---|
| Licence identifiers | MIT + commercial add-ons | GitHub repositoryfetched 2026-10-13 |
| GitHub stars | 35,742 | GitHub repositoryfetched 2026-10-13 |
| Contributors | 221 | GitHub repositoryfetched 2026-10-13 |
| Last update to the code | 2026-10-13 | GitHub repositoryfetched 2026-10-13 |
| Commits by month | 2025-11: 313 · 2025-12: 247 · 2026-01: 179 · 2026-02: 222 · 2026-03: 256 · 2026-04: 241 · 2026-05: 251 · 2026-06: 488 · 2026-07: 676 · 2026-08: 643 · 2026-09: 748 · 2026-10: 350 | awesome-selfhosted listing (counted from GitHub)fetched 2026-10-13 |
| Archived | No | GitHub repositoryfetched 2026-10-13 |
| Current release | v4.57.0 (2026-10-12) | awesome-selfhosted listingfetched 2026-10-13 |
| Docker Hub image | langfuse/langfuse: 19,319,378 pulls, updated 2026-10-13 | Docker Hubfetched 2026-10-13 |
| Platforms | Docker | awesome-selfhosted listingfetched 2026-10-13 |
| External services it needs | clickhouse, minio, redis | docker-compose filefetched 2026-10-13 |
| Listing description | LLM engineering platform for model tracing, prompt management, and application evaluation. Langfuse helps teams collaboratively debug, analyze, and iterate on their LLM applications such as chatbots or AI agents. | awesome-selfhosted listingfetched 2026-10-13 |
| Website | https://langfuse.com | awesome-selfhosted listingfetched 2026-10-13 |
Not on record: Connectors in its directory, API documentation, Single sign-on documentation, Described as an alternative to.
Licence class is derived from the licence identifiers above, never typed by hand. How
Editorial notes
- What it does
- Records the requests your AI app sends to its models and the answers it gets back, so a team can debug problems, manage prompts and rate answer quality.
- Drafted by AI from GitHub repository, reviewed by a person on 2026-10-13
- Goals it fits
- Monitor your AI app's model calls
- Drafted by AI from GitHub repository, reviewed by a person on 2026-10-13
- Ops burden
- 3 of 3 (a cluster or several services)
- Drafted by AI from GitHub repository, reviewed by a person on 2026-10-13
- Known gaps
- Your app needs code changes to send it data. The recommended production setup runs on a Kubernetes cluster. The default Docker setup does not rotate logs, which can fill the disk unless configured.
- Drafted by AI from GitHub repository, reviewed by a person on 2026-10-13
Editorial, not measured. Every note was reviewed by BenchLytix before this record was published.