Query Traces
Langfuse is open-source and data tracked with Langfuse is open. You can query/export data via: SDKs, API, and Langfuse UI.
Common use cases:
- Train or fine-tune models on the production traces in Langfuse. E.g. to create a small model after having used a large model in production for a specific use case.
- Collect few-shot examples to improve quality of output.
- Programmatically create datasets.
Want to learn more about the Langfuse data model? Check out our introduction to tracing in Langfuse.
SDKs
Via the SDKs for Python and JS/TS you can easily query the API without having to write the HTTP requests yourself.
pip install langfusefrom langfuse import Langfuse
langfuse = Langfuse(
secret_key="sk-lf-...",
public_key="pk-lf-...",
host="https://cloud.langfuse.com" # 🇪🇺 EU region
# host="https://us.cloud.langfuse.com" # 🇺🇸 US region
)
# Fetch list of traces, supports filters and pagination
traces = langfuse.fetch_traces()
# Fetch a single trace by ID
trace = langfuse.fetch_trace("traceId")
# Fetch list of observations, supports filters and pagination
observations = langfuse.fetch_observations()
# Fetch a single observation by ID
observation = langfuse.fetch_observation("observationId")
# Fetch list of sessions
sessions = langfuse.fetch_sessions()Python SDK reference including all available filters:
API
All data in Langfuse is available via the API. Refer to the API reference for more information.
Example routes:
GET /api/public/traces
GET /api/public/traces/:traceId
GET /api/public/observations
GET /api/public/observations/:observationId
GET /api/public/sessions
GET /api/public/scores
GET /api/public/scores/:scoreIdExport via UI
All filters applied to the table will be applied to the export.
Available export formats:
- CSV
- JSON
Trouble exporting?
If you’re having trouble exporting data, please let us know and we’ll help you out.