
Open source AI engineering platform. Now part of @clickhousedb. We're hiring: https://teapaw.org/t.co/k6dgv4dws2
github.com/langfuse/langfuse
Joined July 2023
- Tweets1.1K
- Following682
- Followers6.9K
- Likes4K
Pinned Tweet
Replying to @marcklingen
@marcklingen sat down to demo all langfuse features in a single video.
Agent observability, monitoring, evals, prompt management, and what is special about Langfuse.
Great place to start if you are new, or catch up on latest changes.
langfuse.com/watch-demo
What @lotte_verheyden said 👇
Main highlights from our ships this week:
- AI gateway: 6 new providers, 2 new endpoints, budget limits
- Bedrock with IAM roles
- read-access audit logs
- API key presets
- Halo in shortcuts menu
- numeric metadata filters
- saved views for sessions
- J and K keys for annotation queues
- Alerts: tread no data as OK
- Gemini decision models
- default score configs
We are shipping so much, and not talking enough about it.
So I'll start with a video series walking you through our weekly ships at @langfuse. The long list of ships is insane and this is a very concise summary of it. Very excited about this!
1
14
975
We're running through Vegas! 🇺🇸
The @langfuse and @ClickHouseDB team is traveling to Las Vegas for AWS re:invent and we're excited to host our fourth community run after running through Paris, New York City, and Berlin in the last weeks. 5k relaxed pace, open to anyone.
Sign up in the comments if you're around: luma.com/fjoz7ymk 🏃
you might have noticed Langfuse UI has changed dramatically lately. our first designer, @nmtrang29, joined after three years of engineers building Langfuse without one.
wondering what design is like inside an AI-native team? read her reflections: langfuse.com/blog/2026-11-19…
Next up in the 𝗟𝗮𝗻𝗴𝗳𝘂𝘀𝗲 𝗔𝗜 𝗚𝗮𝘁𝗲𝘄𝗮𝘆: @googlecloud Vertex AI.
Use Gemini and open models Google hosts, such as Llama, DeepSeek, Qwen, and gpt-oss, traced in Langfuse. Claude on Vertex AI isn't available yet.
Try the gateway today. Visit langfuse.com/gateway
proactive agents decide on their own whether, when and how far to act
when building them, you're tuning a dial between annoying, risky, and useless
learn how to evaluate when your agent should act, and when it shouldn’t. by @annabellschfr
4
5
391
🚨 Hackathon Announcement
On December 1, @langfuse and Kuro are hosting the Blueprint Hack in Berlin, an evening around building AI for construction.
Construction runs on messy visual data, and that makes it one of the hardest (and most fun) problems to throw AI at.
You’ll learn the language of construction, turn real drawings into real numbers, and compete for €2,000 in AI credits. Beer, pizza and tokens are on us.
There are 30 spots.
Building search on embeddings?
The 𝗟𝗮𝗻𝗴𝗳𝘂𝘀𝗲 𝗔𝗜 𝗚𝗮𝘁𝗲𝘄𝗮𝘆 now serves the OpenAI Embeddings API, for OpenAI, @OpenRouter, and @vercel connections. Same gateway as your other model calls.
Try the gateway today. Visit langfuse.com/gateway
The 𝗟𝗮𝗻𝗴𝗳𝘂𝘀𝗲 𝗔𝗜 𝗚𝗮𝘁𝗲𝘄𝗮𝘆 now serves the @OpenAI Decisions API, for OpenAI and @vercel connections. About 10x faster than Responses, with every call traced as a generation in Langfuse.
Try the gateway today. Visit langfuse.com/gateway
3
20
997
very very neat! the trace transcript which powers this will be used going forward in many parts of langfuse.
through the "normalizedIO" we can deduplicate trace data within a session, to cleanly show each turn.
the new session view was born when trying to use langfuse to improve our agent 😇 . we realized reviewing sessions for chat agents needs a better, chat like interface. often, you can already tell at first glance if a session is good just by the tool called.
we built all of this in!
expect to see the session view in many places in langfuse, soon!
sessions in langfuse now read as one continuous conversation.
under the hood, each trace is turned into a transcript, so user and assistant messages line up across traces. a sidebar jumps between messages, traces, and tool calls.
langfuse.com/changelog/2026-…
langfuse.com retweeted
we redesigned the session view in langfuse. great for whole multi-turn conversations/tasks, you need a lot less clicks now :)
Grok, traced in Langfuse.
@SpaceXSI models are now supported in the 𝗟𝗮𝗻𝗴𝗳𝘂𝘀𝗲 𝗔𝗜 𝗚𝗮𝘁𝗲𝘄𝗮𝘆: use SpaceX Super Intelligence's flagship model for coding and knowledge work, with every call traced natively.
Try the gateway today. Visit langfuse.com/gateway
sessions in langfuse now read as one continuous conversation.
under the hood, each trace is turned into a transcript, so user and assistant messages line up across traces. a sidebar jumps between messages, traces, and tool calls.
langfuse.com/changelog/2026-…
langfuse.com retweeted
Langfuse AI Gateway is easily the most consequential release of the year 👏
Every company needs a reliable way to not only understand, but govern (!) all AI traffic.
Open source Langfuse is already the most widely used platform for AI observability & evals, and now we enter the gateway market. Huge kudos to the team!
Try it out today: langfuse.com/gateway
langfuse.com retweeted
Replying to @langfuse
@langfuse town hall has started!
Community Townhall on Thu
- discuss Gateway
- other major updates to Langfuse
- and roadmap for the coming weeks
luma.com/lvp5fx33
(tune in live to ask questions, will record and upload to youtube/X)
Set up guide here -> langfuse.com/docs/gateway/ge…
197















