langfuse.com (@langfuse) header
langfuse.com (@langfuse) profile picture

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
Photo by langfuse.com (@langfuse)Photo by langfuse.com (@langfuse)Photo by langfuse.com (@langfuse)Photo by langfuse.com (@langfuse)Photo by langfuse.com (@langfuse)Photo by langfuse.com (@langfuse)Photo by langfuse.com (@langfuse)Photo by langfuse.com (@langfuse)Photo by langfuse.com (@langfuse)Photo by langfuse.com (@langfuse)
langfuse.com retweeted
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-…
4
10
605
langfuse.com retweeted
we redesigned the session view in langfuse. great for whole multi-turn conversations/tasks, you need a lot less clicks now :)
1
3
39
1,490
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
Today, we are launching Langfuse AI Gateway, in public beta. We believe every company needs a unified way to use models, track costs and govern access to AI. Change one base_URL in front of OpenAI and Anthropic, and every request becomes a Langfuse trace. 🧵
2
8
27
859
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)
1
10
351
langfuse.com retweeted
you know the team is strong when the releases look like this multiple times A DAY
1
2
14
639
langfuse.com retweeted
Join me for a live lecture on this tomorrow 10am PT! thegenacademy.com/lesson/opt… Editing the last slides, really enjoying how simple and clear some of them are
Wrote about optimizing the cost of your AI agent. This is such an important aspect of building agents successfully, and people often do this "vibes-based". This should offer some structure for anyone going through the process.
3
4
1
34
1,877
RT @marliessophie: We 100x our team in 2 years. > Then: @langfuse offsite 2024 with 7 people > Now: @ClickHouseDB offsite 2026 with 800 p…
1
82
langfuse.com retweeted
Ask Halo challenge The brief: find out why eligible customer applications were getting rejected. In seven minutes, @langfuse's Halo: - found an ignored €6,000 subsidy in traces - tested a revised prompt and - made it live, fixing the application workflow Here's the demo.
3
1
12
880
langfuse.com retweeted
SubQ Snap is now available in @langfuse. Model ID: subq-snap-preview langfuse.com/changelog/2026-…
6
2
38
3,684
langfuse.com retweeted
Here you see Halo help me get started with evals. It looks at my project, determines that the best first step is capturing user feedback signals, and proposes a list of signals I could capture. Really helpful when starting out, Halo can be a great sparring partner in getting to a good observability setup
meet Halo, the assistant that lives inside Langfuse 👀 ask about your traces, sessions and metrics in plain language, or have it build a dataset or dashboard for you. it asks before changing anything.
3
3
2
20
2,231
langfuse.com retweeted
You can also access jev via @OpenRouter now on @langfuse to power your evals
Jev by @typesafeai is now on OpenRouter, in beta. Jev is a System One model. Instead of generating text, it takes your app's state plus a typed question and returns a typed decision with a probability attached. There is no JSON prompting, parsing layer, and nothing to validate against.
5
1
8
1,229
langfuse.com retweeted
Replying to @doneyli @typesafeai
with Jev it is less eval -> eval -> eval and more -> eval -> eval -> eval
2
3
1
32
8,555
langfuse.com retweeted
organizing a chill 5k run with Outrun and our friends at @langfuse before AI Engineer Paris next week. > Thursday, September 24, 7am > Harvest (11e) > Breakfast and coffee after (on us 👀) Sign up here: luma.com/langfuse-umkb @saishrane_
3
6
337
Props to @langfuse for their excellent docs and learning materials! 👏 If you’re trying to wrap your head around concepts like LLM observability, monitoring, and evaluations, these resources are well worth checking out: 🎓 Academy: langfuse.com/academy 🛠️ Workshop: langfuse.com/workshop 📚 Guides: langfuse.com/guides
1
5
14
362
langfuse.com retweeted
New to Langfuse? Try langfuse.com/workshop to explore all capabilities based on a funny example application It is really good now thanks to @annabellschfr!
1
6
730
langfuse.com retweeted
Singapore, our next meetup is headlined by Max Deichmann, Co-founder & CTO of Langfuse 🎤 Building AI Products 🗓️ Friday, Sept 25 · 6:00 PM · AWS Office | Singapore Langfuse is the open-source LLM engineering platform used by thousands of teams to trace, evaluate, and debug AI applications — and it runs on ClickHouse under the hood. Max's talk, "Trace It, Store It, Ship It," goes past the demo stage into what production actually demands. Seats are limited and registration requires approval: buff.ly/zLpT5ZQ Want to speak at a future APJ meetup? buff.ly/ak4tJO8 #ClickHouse #Langfuse #Singapore #LLM #AIEngineering #Observability
2
7
1,288
langfuse.com retweeted
I resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives. More thoughts below.
20,327
166,508
39,480
798,936
175,831,757
langfuse.com retweeted
Replying to @langfuse
3
1
8
2,399
langfuse.com retweeted
One thing I’ve learned from building memory systems: memory is easy to describe at a high level, but surprisingly hard to inspect in practice. That’s why the EverOS integration with @langfuse has been so valuable. It makes the full memory lifecycle visible: what gets stored, what gets recalled, how conflicts are resolved, and what each step costs. That visibility helps users understand the system, while helping us see exactly where EverOS can improve next. Big thanks to the @Langfuse team. Excellent product, and a genuine pleasure to work with.
Memory is the one layer of the agent stack observability can't see. EverOS speaks OTel natively now, so memory operation land in @langfuse: what got stored, what got recalled, how confident that recall was, how conflicting memories got merged, and what it cost. Stop guessing why your agent forgot: langfuse.com/integrations/ot…
1
1
7
1,733
langfuse.com retweeted
Evaluators and datasets are very closely linked, great write-up by @annabellschfr!
7
65
11,325