Braintrust vs Langfuse
Prices and features checked 2026-10-17 on both vendors' own pages. “not verified” means we could not confirm it.
What is the difference between Braintrust and Langfuse? Langfuse is open source (MIT core) and free to self-host, and its cloud bills by units (traces, observations and scores) from $29/month on Core. Braintrust is a hosted platform, with self-hosting on Enterprise, billed by processed data and scores, with a $0 Starter plan and Pro at $249/month. Both offer OpenTelemetry ingestion, CI runs, LLM-as-judge scoring, human review, prompt versioning and dataset versions.
| Braintrust | Langfuse | |
|---|---|---|
| Made by | Braintrust Data | Langfuse (ClickHouse, Inc.) |
| Licence | Hosted service; its autoevals scorer library is MIT [source] | MIT for code outside the ee/ directories; ee/ code under a separate licence [source] |
| Self-hosting | Self-hosted data plane; on-prem or hosted deployment on Enterprise [source] | Free to self-host (Docker Compose or Kubernetes) [source] |
| Meter | Processed data (GB) plus scores, plus model credits at token rates [source] | Billable units: traces, observations (spans, events, generations) and scores [source] |
| Free tier | Starter $0: 1 GB processed data, 10k scores, 14-day retention, unlimited users; then $4/GB and $2.50 per 1k scores [source, checked 2026-10-17] | Hobby: 50k units/month, 30 days data access, 2 users [source, checked 2026-10-17] |
| Entry paid price | Pro $249/month: 5 GB and 50k scores included, then $3/GB and $1.50 per 1k scores; 30-day retention [source, checked 2026-10-17] | Core $29/month: 100k units included, then $8 per 100k units (graduated down to $6 per 100k above 50M) [source, checked 2026-10-17] |
| Retention | 14 days (Starter), 30 days (Pro), up to 180 days at $0.50/GB/month [source] | Data access: 30 days (Hobby), 90 days (Core), 3 years (Pro and Enterprise) [source] |
| OpenTelemetry | Accepts OpenTelemetry traces and logs [source] | OpenTelemetry ingestion on all plans [source] |
| CI gating | GitHub Actions or the bt eval CLI [source] | Run experiments in CI and gate changes [source] |
| LLM-as-judge | LLM-as-a-judge and code scorers [source] | LLM-as-judge evaluators on all plans [source] |
| Human annotation | Human review scores on traces and experiments [source] | Annotation queues: 1 on Hobby, 3 on Core, unlimited on Pro [source] |
| Prompt management | Prompt version history, restore and compare [source] | Prompt versioning and release labels [source] |
| Dataset versioning | Pin dataset versions with snapshots [source] | Dataset versions are accessible via the SDK/API [source] |
| SOC 2 / HIPAA | SOC 2 Type II listed; BAA for HIPAA on Enterprise [source] | SOC 2 Type II and ISO 27001 reports and HIPAA (BAA) on Pro ($199/month) and Enterprise [source] |
Verdict: when to use which
Use Braintrust if you want a hosted tool where the free plan already covers unlimited users and small volumes, and your usage is easier to estimate as processed GB and number of scores than as trace units.
Use Langfuse if you want to run open-source code on your own infrastructure at no licence cost, or want a low flat cloud fee ($29/month Core) with unlimited users and per-unit usage.
The meters differ, so price both on a week of your own traffic in the pricing calculator.
FAQ
Is Langfuse cheaper than Braintrust?
It depends on volume and team size. Langfuse's own example: 200k units a month on Core costs $37 ($29 plus $8 for the second 100k units). Braintrust Starter is $0 up to 1 GB of processed data and 10k scores, then $4/GB and $2.50 per 1k scores. Try your own numbers in our pricing calculator.
Can I self-host Braintrust or Langfuse?
Langfuse: yes, the open-source version is free to self-host. Braintrust: a self-hosted data plane is documented, and its pricing page lists on-prem or hosted deployment on Enterprise.
Which has HIPAA support?
Langfuse lists HIPAA (BAA) on Pro and Enterprise. Braintrust lists a BAA on Enterprise plans.
More
Other tool comparisons: Langfuse vs LangSmith · Braintrust vs LangSmith · Arize Phoenix vs Langfuse · Comet Opik vs Langfuse · Promptfoo vs DeepEval / Confident AI