Metadata

Overview

With Confident AI, you can attach additional metadata to traces, spans, and threads. This information can be used for filtering, grouping, and analyzing your traces in the observatory.

Add Metadata

Use the update current span and update current trace functions to add span-level and trace-level metadata.

Each metadata argument is a dictionary whose keys are strings and whose values are any JSON-serializable type.

1
from deepeval.tracing import observe, update_current_span, update_current_trace
2
3
@observe()
4
async def llm_app(query: str):
5
    # Add span-level metadata
6
    update_current_span(
7
        metadata={
8
            "source": "knowledge_base_1",
9
            "retrieved_documents": 3
10
        }
11
    )
12
    
13
    # Add trace-level metadata
14
    update_current_trace(
15
        metadata={
16
            "user_id": "user-456",
17
            "app_version": "1.2.3",
18
        }
19
    )
20
21
llm_app("Test Metadata")

Thread-Level Metadata

You can also attach metadata to threads — useful for tagging production conversations with attributes like DVA version, client, or agent ID. See Set Thread Metadata for the payload shape.