Project metadata
Names, structure, and relationships for each Project.
Prompt metadata
A managed prompt in Langfuse Prompt Management — the literal template that runs in production, plus its model config, deployment labels, tags, and version history.
Dataset metadata
A Langfuse evaluation dataset — the fixed set of inputs and expected outputs a model or prompt change is measured against, plus its run history.
Score Definition metadata
A Langfuse score configuration — the rubric that gives a recorded score value its meaning: data type, numeric bounds, and named categories.
Trace Type metadata
One named kind of traced operation in this Langfuse project, with its observed volume, latency, cost and error profile.
Score Type metadata
One distinct score name actually recorded in this Langfuse project, with its observed volume and value distribution.