Source · release
Gemini API changelog, 2026-11-05: Omni 1.2 Flash GA, richer function responses
Summary
The Gemini API changelog for 2026-11-05 has three changes. Tools can now return Office and OpenDocument files, audio, and video to the model in function responses; the API accepts more document formats as input; and the video model Gemini Omni 1.2 Flash is generally available, with Omni 1.1 Flash requests redirected to it from 2026-12-03.
Key points
- Richer function responses: a tool result can include Office or OpenDocument files, audio, or video (inline or as file URIs), in both the GenerateContent and Interactions APIs. An agent's tool can hand back, say, a generated spreadsheet or a recording for the model to read directly.
- More input formats: PPT, ODP, and ODS are accepted. DOC, PPT, and ODP are processed as page images; for ODS the model receives the data and formulas of the first sheet.
- Gemini Omni 1.2 Flash GA (
gemini-omni-1.2-flash): video up to 15 seconds, conversational editing, better subject consistency across up to six reference images, 30% lower video output price, and doubled paid-tier rate limits. - Migration: from 2026-12-03, calls to
gemini-omni-1.1-flashare routed to 1.2 Flash and billed at its rates.
Why it matters
Multimodal tool results remove a common workaround (upload a file, then reference it) when agent tools produce documents or media.