●Change8

Migrating to Datadog SDK v4.13.0rc1

Version v4.13.0rc1 introduces 1 breaking change. This guide details how to update your code.

Released: 7/27/2026

1
Breaking Changes
10
Migration Steps
19
Affected Symbols

⚠️ Check Your Code

If you use any of these symbols, you need to read this guide:

ddtrace.appsec.ai_guardddtrace.aiguardddtrace.aiguard.integrations.strandsddtrace.DDTraceDeprecationWarningddtrace/contrib/internal/pytestddtrace/testing/internal/pytestLLMObs.get_prompt()LLMObs.refresh_prompt()LLMObs.enable()span constructorsdecoratorssubmit_evaluation()get_spans()publish_evaluator()OpenAI integrationclient.realtime.connect()mistralai clientOpenTelemetryAWS Lambda MicroVM

Breaking Changes

●Issue #1

The `ml_app` argument and `DD_LLMOBS_ML_APP` environment variable for LLM Observability are deprecated and will be removed in the next major version. Use `agent_service` and `DD_SERVICE` instead to group traces belonging to a root agentic service.

Migration Steps

  1. 1
    Update imports from `ddtrace.appsec.ai_guard` to `ddtrace.aiguard`.
  2. 2
    For Strands Agents classes, import them from `ddtrace.aiguard.integrations.strands`.
  3. 3
    Remove the `DD_PYTEST_USE_NEW_PLUGIN` environment variable from your configuration.
  4. 4
    Update imports such as `from ddtrace.appsec.ai_guard import AIGuardClient` to `from ddtrace.aiguard import AIGuardClient`.
  5. 5
    Set `DD_ENV` instead of using the `label` parameter for `LLMObs.get_prompt()` and `LLMObs.refresh_prompt()`.
  6. 6
    Use `agent_service` and `DD_SERVICE` instead of the `ml_app` argument and `DD_LLMOBS_ML_APP` environment variable for LLM Observability.
  7. 7
    When using Bedrock calls with an application-inference-profile ARN as `modelId`, ensure `DD_LLMOBS_ENABLED=true` and you have `bedrock:GetInferenceProfile` IAM permission if you want to resolve the underlying foundation model.
  8. 8
    Set `DD_ENV` to select the prompt version for a specific environment when using `LLMObs.get_prompt()`.
  9. 9
    Set `DD_OPENAI_REALTIME_ENABLED=false` to disable Realtime API instrumentation for the OpenAI integration.
  10. 10
    Set `DD_EXPERIMENTAL_FLAGGING_PROVIDER_SPAN_ENRICHMENT_ENABLED=true` to enable experimental APM span enrichment for the OpenFeature provider.

Release Summary

This release introduces significant enhancements to LLM Observability, including prompt management, image attachment, and OpenAI Realtime API instrumentation. It also brings improvements to tracing safety, OpenTelemetry integration, AI Guard features, and crash tracking.

Need More Details?

View the full release notes and all changes for Datadog SDK v4.13.0rc1.

View Full Changelog