Langfuse is an open source LLM engineering platform that helps teams manage prompts, trace and debug AI agents, and run evaluations. You can self-host it for free or use Langfuse Cloud, which has a free Hobby plan and paid plans from $29 a month. Langfuse is now part of ClickHouse.
What is Langfuse?
Langfuse is one of the most popular open source tools for building production AI applications, with more than 36,000 GitHub stars. It gives developers a central place to version and deploy prompts, see every step of an LLM call or agent run, and measure quality with evaluations.
Its prompt management lets teams edit prompts in a UI, test them in a playground, and fetch the latest version in code without redeploying. Tracing shows inputs, outputs, latency and cost for every request, while datasets, LLM-as-judge evaluators and human annotation help improve quality. Langfuse integrates with OpenAI, LangChain, LlamaIndex, Vercel AI SDK and many more.
Langfuse key features
Prompt management
Version, label and deploy prompts without code changes.
Prompt playground
Test prompts and models side by side.
Tracing
See every step of LLM calls and agent runs.
Cost and latency tracking
Monitor token usage, cost and speed.
Evaluations
Run LLM-as-judge and custom evals on outputs.
Datasets and experiments
Test prompt changes against benchmark datasets.
Human annotation
Review and score outputs in annotation queues.
Open source
Self-host for free with Docker or Kubernetes.
Integrations
Works with OpenAI, LangChain, LlamaIndex and more.
Langfuse pricing plans (2026)
Langfuse can be self-hosted for free. Langfuse Cloud has a free Hobby plan and paid plans billed monthly, with usage-based pricing for extra units.
Hobby
50k units a month, 30 days data access, 2 users and community support.
Core
100k units a month, 90 days data access, unlimited users and in-app support.
Pro
100k units a month, 3 years data access, unlimited annotation queues and priority support.
Enterprise
Everything in Pro plus a named lead engineer, support SLA and custom terms.
Extra usage costs $8 per 100k units. Self-hosted enterprise features are priced separately. Check the latest prices on the official Langfuse pricing page.
How to use Langfuse
- 1
Go to langfuse.com and sign up for Langfuse Cloud, or self-host it.
- 2
Create a project and copy your API keys.
- 3
Add the Langfuse SDK or integration to your app.
- 4
Move your prompts into Langfuse prompt management.
- 5
Review traces and set up evaluations to improve quality.
Who should use Langfuse?
AI engineers
Debug and monitor LLM apps in production.
Prompt engineers
Version and test prompts collaboratively.
Product teams
Track quality, cost and user feedback.
Enterprises
Self-host for data control and compliance.
Pros and cons
What we like
- Open source and self-hostable
- Prompt management and tracing in one tool
- Generous free Hobby plan
- Many framework integrations
Watch out for
- Built for developers, not beginners
- Usage costs grow at scale
- Short data retention on Hobby
- Advanced security needs higher plans
Best Langfuse alternatives
Langfuse is one of the best tools for prompt management and LLM observability, especially for teams that value open source. Free self-hosting and a generous Hobby plan make it easy to start, and Core is affordable for small teams in production.
Langfuse FAQs
Is Langfuse free?
Yes. Langfuse is open source and free to self-host, and Langfuse Cloud has a free Hobby plan.
How much does Langfuse cost?
Langfuse Cloud Core costs $29 a month, Pro $199 and Enterprise $2,499, plus $8 per extra 100k units.
What is prompt management in Langfuse?
It lets you version prompts, edit them in a UI and fetch them in code without redeploying.
Can I self-host Langfuse?
Yes. You can self-host Langfuse with Docker Compose, Kubernetes or Terraform.
Who owns Langfuse?
Langfuse has joined ClickHouse and now appears as Langfuse by ClickHouse.
Does Langfuse work with LangChain?
Yes. Langfuse integrates with LangChain, LlamaIndex, OpenAI SDKs and many more.
