AI Devtools Radar

Langfuse

LLM observabilitylangfuse.com15 published changes

2 sources under watch

Change timeline

deprecation

Gemini 3 Pro/3.1 Flash-Lite preview models flagged for shutdown

Users relying on Gemini 3 Pro and 3.1 Flash-Lite preview models will need to migrate to alternative models as these are being deprecated.

Evidence
+chore(pricing): flag shut-down Gemini 3 Pro/3.1 Flash-Lite previews, refresh audit snapshot

github.com

otherhigh

Security fix: JWT session max age reduced to 14 days

All users will have sessions automatically expire after 14 days instead of previous default, requiring re-authentication for longer-lived sessions.

Evidence
+security(auth)!: reduce default JWT session max age to 14 days

github.com

feature

OIDC profile can now be read from the userinfo endpoint

Users with custom OIDC configurations can have their profile information automatically read from the userinfo endpoint, improving authentication flexibility.

Evidence
+feat(auth): read custom OIDC profile from the userinfo endpoint

github.com

other

Fixed surrogate pair splitting in truncate() function

Text truncation now correctly preserves Unicode surrogate pairs, preventing character corruption when truncating strings with emoji or other multi-byte characters.

Evidence
+fix(shared): truncate() no longer splits surrogate pairs

github.com

api

Add model definition upsert endpoint to API

Developers can now use a new API endpoint to upsert model definitions, improving model management capabilities.

Evidence
+feat(api): add model definition upsert endpoint by @hassiebp in #16516

github.com

deprecationhigh

API v3 sunset date announced for November 16, 2026

Users of Langfuse API v3 need to plan migration to v4 before November 16, 2026 as v3 will be sunset on that date.

Evidence
+feat(api): state the concrete v3 sunset date, November 16, 2026 (LFE-10895)

github.com

feature

OpenAI Responses API request input now renders in trace chat view

Users can now see OpenAI Responses API request details in the trace chat view, improving debugging and monitoring capabilities.

Evidence
+fix(chatml): render OpenAI Responses API request input in trace chat view

github.com · also seen at 1 more source

positioning

Renamed 'Monitors and Alerts' feature to 'Alerts' across all pricing tiers

Users and customers will see the feature referred to as 'Alerts' instead of 'Monitors and Alerts', potentially affecting how they understand the feature offering and documentation references.

Evidence
−Monitors and Alerts 2 monitors Monitors and Alerts 20 monitors Monitors and Alerts 50 monitors Monitors and Alerts 100 monitors
+Alerts 2 alerts Alerts 20 alerts Alerts 50 alerts Alerts 100 alerts

langfuse.com

other

Fix unicode escaping in client-side trace downloads and copies

Users can now correctly download and copy traces with unicode characters without escape sequence corruption.

Evidence
+fix(trace): decode unicode escapes in client-side trace downloads and copies

github.com

pricing

Added GPT-5.6 usage aliases for pricing calculations

GPT-5.6 model pricing is now properly configured with usage aliases.

Evidence
+fix(pricing): add GPT-5.6 usage aliases

github.com

apihigh

Observations API v2 and Metrics API v2 released with improved performance

Developers using the API get significantly faster performance with v2 endpoints; migration from v1 recommended.

Evidence
+the significantly faster Observations API v2 and Metrics API v2

github.com

feature

Performance telemetry added for lazy JSON viewer in trace view

Developers will have better observability into trace JSON viewer performance metrics.

Evidence
+feat(trace): perf telemetry for the lazy JSON viewer (LFE-14419)

github.com · also seen at 1 more source

other

Media retention lease renewal during blob cleanup

Prevents potential data loss by ensuring blob cleanup operations properly maintain retention leases.

Evidence
+fix: renew media retention leases during blob cleanup

github.com

deprecationhigh

Superseded Postgres and ClickHouse tables dropped in v4

Users upgrading to v4.0.0 must ensure they have migrated away from legacy table schemas before upgrading, as deprecated tables are removed.

Evidence
+feat(v4)!: drop superseded Postgres and ClickHouse tables

github.com

feature

Added Observations API v2 as a new rate-limited API endpoint across all pricing tiers

Developers now have explicit rate limits for the new Observations API v2 (30-1000 requests/min depending on tier), providing clarity on API usage constraints for this new endpoint.

Evidence
−Rate limit (general API routes) Rate limit (datasets api) Rate limit (legacy Metrics API)
+Rate limit (General API) Rate limit (Datasets API) Rate limit (Observations API v2) Rate limit (Legacy Metrics API)

langfuse.com · also seen at 1 more source