compare
Langfuse vs OpenLIT
The same facts for both, read from GitHub every night, and the relation a person reviewed.
| Fact | Langfuse | OpenLIT |
|---|---|---|
| Language | TypeScript | TypeScript |
| Licence | Other | Apache-2.0 |
| Stars | 35k | 2.8k |
| Latest | v4.53.0 | openlit-2.1.0 |
| Last push | 2026-10-06 | 2026-10-05 |
| Release cadence | about 1 day between releases | about 8 days between releases |
| Active contributors | 33+ commit authors on the default branch in the last 90 days | 46 commit authors on the default branch in the last 90 days |
| Flags | none | none |
How they relate
Both replace LangSmith. Alternatives to LangSmith →
Full replacementLangfuseTracing, prompt management, datasets and evaluations, with OpenTelemetry ingestion.
PartialOpenLITOpenTelemetry-native traces and metrics for LLM apps, GPUs and vector databases, with prompt management.
Langfuse
- v4.53.02026-10-06feat(ai-gateway): expose resolution context in response headers, spans and logs
- v4.52.02026-10-06feat(trace-batch): bound batches by estimated weight and exclude oversized traces
- v4.51.02026-10-05feat(mcp): support decision-model evaluators
- v4.50.02026-10-02feat(web): show error id in tRPC error toast and include it in issue reports
- v4.49.02026-10-01feat(web): add org-level ingestion overview tRPC route
OpenLIT
- openlit-2.1.02026-09-10openlit: 2.1.0
- openlit-2.0.02026-08-28openlit: 2.0.0
- otel-gpu-collector-0.0.82026-08-26otel-gpu-collector: 0.0.8
- ts-1.15.02026-08-03Add per-evaluation-type threshold scores and API-key-authenticated create/update endpoints for offline evaluation types…
- py-1.45.02026-08-03Aligned offline/SDK evaluations with tenant-configured evaluation types including rule-engine context resolution and pe…
Langfuse
unsigned The latest release, v4.53.0, carries no signature GitHub could verify.
Loading the security report
OpenLIT
unsigned The latest release, openlit-2.1.0, carries no signature GitHub could verify.
Loading the security report