Field noteOct 24, 20261 min read1 source

Google ADK for Python v2.13.0 adds classifier-based workflow routing, CascadeLive barge-in voice and MongoDB session/memory services

In one minute

Google released ADK for Python 2.13.0, dated October 22, 2026.

What changed: ADK Python v2.13.0 adds classifier-based workflow routing, CascadeLive barge-in voice, and MongoDB session and memory services; the priors cover the toolkit generally and a separate voice-agent guide.

The release promotes 12 experimental features to stable status and adds classifier components for workflow routing, including ClassifierNode, LlmClassifier, and a TypesafeClassifier integration for TypeSafe Jev models.

CascadeLive gains barge-in support plus Gemini speech-to-text and text-to-speech integration. ADK also adds experimental MongoDbSessionService and MongoDbMemoryService. The session service can share sessions across runner replicas; the memory service recalls past turns by keyword match and does not require an embedding model or vector index. The release notes claim up to 8x faster execution turns for string-model agents.

The upgrade changes telemetry defaults to OpenTelemetry GenAI semantic conventions v2, otel_semconv_1_44, and removes the call_llm span. Users needing the legacy v1 schema can set ADK_TELEMETRY_SCHEMA_VERSION_OPT_IN=otel_semconv_1_36. The source does not independently establish the claimed performance gain or migration impact for individual deployments. Source: https://github.com/google/adk-python/releases

Evidence / sources

  1. 1github.com/google/adk-python/releases
All Field Notes

Next reads

Read next

Open the graph
  1. 01topicGoogleExplore the google topic hub.
  2. 02archiveField NotesOpen the latest editorial analysis.
  3. 03source ledgerSource LedgerInspect the public source evidence graph.
Who read this page?Machine requests, hidden until opened

Loading the privacy-safe route aggregate...

Open the JSON contract