Software product / application
Langfuse Cloud
Trace AI application behavior, manage prompt versions and evaluate outputs in a shared engineering platform.
About this listing
Langfuse Cloud brings tracing, prompt management and evaluation together for teams building AI applications. Traces can capture model calls alongside retrieval, embeddings and other application steps, with sessions for multi-turn conversations and views of latency and cost. Python and JavaScript SDKs, OpenTelemetry and framework integrations provide ways to send application activity into the platform.
Teams can version prompts, deploy them through labels and compare their behavior in linked traces. Evaluation supports production scoring and experiments against datasets, including human feedback and custom evaluation results. This makes the platform useful when a team needs to investigate an agent's execution and test changes against repeatable examples.
The hosted service offers a free Hobby tier with usage and retention limits, alongside paid plans. Model-provider usage is separate from the platform subscription; check the current pricing page for allowances and paid usage terms.