Skip to content

langfuse

v3.180.0 Feature

This release adds 3 notable features for engineering teams evaluating rollout.

Published 1mo Tracing
✓ No known CVEs patched
Read the diff → Tool health → What is this tool? →

✓ No known CVEs patched in this version

Topics

analytics autogen evaluation langchain large-language-models llama-index
+12 more
llm llm-evaluation llm-observability llmops monitoring observability openai playground prompt-engineering prompt-management self-hosted ycombinator

Summary

AI summary

Release v3 scores API, flatten response shape, add GCP storage option, improve agent UI.

Changes in this release

Feature Low

Adds v3 scores API release.

Adds v3 scores API release.

Source: llm_adapter@2026-06-09

Confidence: high

Feature Low

Adds GOOGLE_CLOUD_UNIVERSE_DOMAIN config for GCP blob storage.

Adds GOOGLE_CLOUD_UNIVERSE_DOMAIN config for GCP blob storage.

Source: llm_adapter@2026-06-09

Confidence: high

Feature Low

Adds metric alerting (scheduler, processor, dispatch UI) to monitors.

Adds metric alerting (scheduler, processor, dispatch UI) to monitors.

Source: llm_adapter@2026-06-09

Confidence: high

Feature Low

Adds code evaluators to public API endpoints.

Adds code evaluators to public API endpoints.

Source: llm_adapter@2026-06-09

Confidence: high

Bugfix Medium

Suppresses network connectivity error toasts.

Suppresses network connectivity error toasts.

Source: llm_adapter@2026-06-09

Confidence: high

Bugfix Medium

Passes SMTP_CONNECTION_URL through NextAuth instead of SES options.

Passes SMTP_CONNECTION_URL through NextAuth instead of SES options.

Source: llm_adapter@2026-06-09

Confidence: high

Bugfix Medium

Limits 'hasAllTokens' function to a maximum of 64 search tokens.

Limits 'hasAllTokens' function to a maximum of 64 search tokens.

Source: llm_adapter@2026-06-09

Confidence: high

Bugfix Medium

Preserves unsafe dataset integers in evals.

Preserves unsafe dataset integers in evals.

Source: llm_adapter@2026-06-09

Confidence: high

Bugfix Medium

Returns 503 for public auth DB failures.

Returns 503 for public auth DB failures.

Source: llm_adapter@2026-06-09

Confidence: low

Refactor Low

Removes foreign keys from project-media.

Removes foreign keys from project-media.

Source: llm_adapter@2026-06-09

Confidence: high

Full changelog

What's Changed

Features

feat(agent): Trace in-app agent in langfuse by @bezbac in https://github.com/langfuse/langfuse/pull/13748
feat(scores): version v3 cursor with v: 1 for forward migration by @niklassemmler in https://github.com/langfuse/langfuse/pull/14084
feat(scores): flatten v3 response shape to match proposal by @niklassemmler in https://github.com/langfuse/langfuse/pull/14088
feat(scores): accept lowercase source and dataType on v3 list endpoint by @niklassemmler in https://github.com/langfuse/langfuse/pull/14092
feat(scores): reject unknown query params on v3 list endpoint by @niklassemmler in https://github.com/langfuse/langfuse/pull/14093
feat(agent): Show tool calls in the UI by @bezbac in https://github.com/langfuse/langfuse/pull/14086
feat(evals): allowlist more TypeScript evaluator helpers by @wochinge in https://github.com/langfuse/langfuse/pull/14097
feat(scores): release v3 scores API by @niklassemmler in https://github.com/langfuse/langfuse/pull/14099
feat(storage): add GOOGLE_CLOUD_UNIVERSE_DOMAIN configuration option for GCP blob storage by @fmontalbano-mgen in https://github.com/langfuse/langfuse/pull/13453
feat(agent): Add Langfuse Docs MCP by @bezbac in https://github.com/langfuse/langfuse/pull/14116
feat(agent): Improve in-app agent UI by @bezbac in https://github.com/langfuse/langfuse/pull/14027
feat(support): Team/Enterprise urgent (Sev-1) support requests by @jannikmaierhoefer in https://github.com/langfuse/langfuse/pull/14094
feat(api): expose code evaluators in public endpoints by @wochinge in https://github.com/langfuse/langfuse/pull/14058
feat(monitors): metric alerting - scheduler, processor, alert dispatch, and management ui by @marksalpeter in https://github.com/langfuse/langfuse/pull/13874
feat(fern/scores): demote scores v2 to legacy, promote v3 to canonical Scores section by @niklassemmler in https://github.com/langfuse/langfuse/pull/14126

Fixes

fix(web): Margins on project overview by @bezbac in https://github.com/langfuse/langfuse/pull/14078
fix: suppress network connectivity error toasts by @maxdeichmann in https://github.com/langfuse/langfuse/pull/14054
fix(api): return 503 for public auth db failures by @peter941221 in https://github.com/langfuse/langfuse/pull/13866
fix(email): pass SMTP_CONNECTION_URL through NextAuth instead of SES options by @6matt in https://github.com/langfuse/langfuse/pull/14100
fix: Function 'hasAllTokens' supports a max of 64 search tokens by @sumerman in https://github.com/langfuse/langfuse/pull/14114
fix(scores): qualify scores columns by @Epochex in https://github.com/langfuse/langfuse/pull/13882
fix(playground): parse pipecat tools by @nimarb in https://github.com/langfuse/langfuse/pull/14120
fix(evals): preserve unsafe dataset integers by @wochinge in https://github.com/langfuse/langfuse/pull/14119

Chores

docs(agents): add security-review skill for SSRF and recurring findings by @Steffen911 in https://github.com/langfuse/langfuse/pull/14087
docs(skills): require valid markdown output by @maxdeichmann in https://github.com/langfuse/langfuse/pull/14091
chore: remove foreign keys from project-media by @maxdeichmann in https://github.com/langfuse/langfuse/pull/14105
chore: Enable eslint concurrency by @bezbac in https://github.com/langfuse/langfuse/pull/14115
docs: preserve README image aspect ratios by @nkabardin in https://github.com/langfuse/langfuse/pull/14131

New Contributors

  • @fmontalbano-mgen made their first contribution in https://github.com/langfuse/langfuse/pull/13453
  • @Epochex made their first contribution in https://github.com/langfuse/langfuse/pull/13882

Full Changelog: https://github.com/langfuse/langfuse/compare/v3.179.1...v3.180.0

Weekly OSS security release digest.

The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.

No spam, unsubscribe anytime.

Share this release

Track langfuse

Get notified when new releases ship.

Sign up free

About langfuse

🪢 Open source LLM engineering platform: LLM Observability, metrics, evals, prompt management, playground, datasets. Integrates with OpenTelemetry, Langchain, OpenAI SDK, LiteLLM, and more. YC W23

All releases →

Related context

Beta — feedback welcome: [email protected]