Discovery API:aiplatform.v1
Medium Impact
Docs

Vertex AI adds Audio Transcription and Memory Event Ingestion

A new set of schemas for Audio Transcription has been added, supporting speaker diarization, word-level timestamps, and language auto-detection. The Memory Bank resource now includes an 'ingestEvents' method, further maturing the agentic memory framework. For video generation, the 'omniRewriter' configuration allows tuning chunk durations and input FPS. Additionally, 'mutateDeployedIndex' now supports an 'updateMask', enabling partial updates for resources like 'dedicated_resources' and 'access_logging'.