ContextQMD
Back to Opik

Observability for Prompt Flow with Opik

apps/opik-documentation/documentation/fern/docs/tracing/integrations/prompt-flow.mdx

2.0.24-52622.6 KB
Original Source

Prompt Flow enables tracing through Prompt Flow APIs and uses OpenTelemetry export settings for outbound telemetry.

When this guide applies

Use this when your Prompt Flow project already uses Prompt Flow tracing and you want its telemetry exported into Opik.

Opik OTLP endpoint modes

For full endpoint/header details, see Opik OpenTelemetry overview.

<Tabs> <Tab value="Opik Cloud" title="Opik Cloud"> ```bash wordWrap export OTEL_EXPORTER_OTLP_ENDPOINT=https://www.comet.com/opik/api/v1/private/otel export OTEL_EXPORTER_OTLP_HEADERS='Authorization=<your-api-key>,Comet-Workspace=<your-workspace>,projectName=<your-project-name>' ``` 
    Required headers: `Authorization`, `Comet-Workspace`
    Optional headers: `projectName`
</Tab>
<Tab value="Enterprise deployment" title="Enterprise deployment">
    ```bash wordWrap
    export OTEL_EXPORTER_OTLP_ENDPOINT=https://<comet-deployment-url>/opik/api/v1/private/otel
    export OTEL_EXPORTER_OTLP_HEADERS='Authorization=<your-api-key>,Comet-Workspace=<your-workspace>,projectName=<your-project-name>'
    ```

    Required headers: `Authorization`, `Comet-Workspace`
    Optional headers: `projectName`
</Tab>
<Tab value="Self-hosted instance" title="Self-hosted instance">
    ```bash wordWrap
    export OTEL_EXPORTER_OTLP_ENDPOINT=http://localhost:5173/api/v1/private/otel
    export OTEL_EXPORTER_OTLP_HEADERS='projectName=<your-project-name>'
    ```

    Required headers: none by default
    Optional headers: `projectName`, auth headers if enabled
</Tab>
</Tabs>

Example (Prompt Flow tracing API)

Intent: Enable Prompt Flow tracing and export telemetry through your configured OTLP endpoint.

Applies when: You use Prompt Flow's native tracing API and want those traces exported via your OTEL settings.

Required inputs:

  • Prompt Flow tracing enabled in code
  • OTEL endpoint/header env vars for your Opik mode

Optional inputs:

  • service/resource metadata
  • additional OTel collector routing or enrichment

Minimal valid setup:

python
from promptflow.tracing import start_trace

start_trace()

# Run prompt flow pipeline as usual.

Validation

  1. Run a flow with start_trace() enabled.
  2. Confirm telemetry export calls are made to your configured OTLP endpoint.
  3. Verify trace records appear in the expected Opik project.

Source references