Gemini Omni 1.2 Flash adds longer video generation and editing
Gemini Omni 1.2 Flash is generally available with longer video generation, conversational editing, reference-image consistency, and a migration path from 1.1.
Editorial Team
The Authority AI Tools editorial team maintains this directory using vendor documentation, dated source checks, product changelogs, and clearly identified hands-on observations where available.
Google’s Gemini API changelog lists two significant late-August general-availability releases: Gemini 3.5 Transcribe on August 26 and Gemini Omni Flash on August 27. The releases target different workloads—low-latency speech-to-text and conversational video generation/editing—but both add production-oriented endpoints for developers.
Gemini 3.5 Transcribe for speech-to-text
The release includes gemini-3.5-transcribe, a dedicated non-streaming speech-to-text model. Google lists utterance-based language detection across more than 85 languages, speaker diarization, word-level timestamps, and custom vocabulary biasing for up to 1,000 terms.
Google also lists gemini-3.5-transcribe-live, a bidirectional streaming speech-to-text model over WebSockets through the Live API. It supports interim and final transcription events, Smart transcription mode, and multiple voice-activity-detection strategies.
These are model capabilities described in the API changelog, not a promise that every language or configuration has identical quality. Teams should test latency, diarization, and vocabulary behavior against their own audio before switching production traffic.
Gemini Omni Flash for video generation and editing
gemini-omni-1.1-flash is now generally available as a fast, conversational video generation and editing model. The API supports extending an existing video with the extend task, and interpolation between two images with the image_to_video task.
The release adds a resolution parameter in video_config with 360p, 720p (the default), 1080p, and 4k outputs. Google notes that 1080p and 4K outputs use upscaling, which is important when evaluating visual quality and processing expectations.
The older gemini-omni-flash-preview endpoint is scheduled for deprecation on September 30, 2026. Integrations using the preview endpoint should plan a migration rather than treating the GA release as a drop-in change without testing.
November 5 follow-up: Gemini Omni 1.2 Flash
Google marked gemini-omni-1.2-flash generally available on November 5. The Gemini API changelog lists video generation up to 15 seconds, first-and-last-frame interpolation and video extension in short turns, and improved character and subject consistency with up to six reference images. Google says video-output pricing is 30% lower at $12 per million tokens and paid-tier rate limits are doubled; verify the current pricing and quota pages for the account and use case. (https://ai.google.dev/gemini-api/docs/changelog; https://ai.google.dev/docs/omni)
Google also says requests to gemini-omni-1.1-flash will automatically route to gemini-omni-1.2-flash starting December 3, 2026, at the 1.2 model’s rates. Update the model string and test output behavior before that date if the application needs to pin a specific version. (https://ai.google.dev/gemini-api/docs/changelog)
Sources
- Google AI for Developers — Gemini API changelog (August 26–27, 2026): https://ai.google.dev/gemini-api/docs/changelog
- Google for Developers on X — official developer updates: https://x.com/googledevs
- Google AI for Developers — Gemini API November 2026 changelog: https://ai.google.dev/gemini-api/docs/changelog
- Google AI for Developers — Gemini Omni Flash guide: https://ai.google.dev/docs/omni
Free Resource
2026 AI Coding Tools Comparison Chart
Side-by-side comparison of features, pricing, and capabilities for every major AI coding tool.
No spam, unsubscribe anytime.
Frequently Asked Questions
What is Gemini Omni 1.2 Flash adds longer video generation and editing?
Related Articles
Cursor Cloud Agents can now build and publish live apps
Cursor Apps lets Cloud Agents build web apps, provision hosting and Postgres, and publish only after required security checks and an explicit user action.
Read more →Product UpdatesOpenAI Images 2.6 adds faster generation, GIFs, and editing controls
OpenAI's GPT-Image-2.6 Sunburst and Flare models add GIF and SVG output, markup-guided editing, and adjustable complexity across ChatGPT, Codex, and the API.
Read more →Product UpdatesOpenAI expands Daybreak Red access for approved security teams
OpenAI announced GPT-6 Cyber for approved Daybreak Red customers and brought GPT-5.6 Cyber to approved AWS GovCloud customers through Amazon Bedrock.
Read more →