Veo 3.1 and Veo 3.1 Fast previews end October 22: what replaces them
Google's deprecations page sets October 22, 2026 as the shutdown date for the Veo 3.1 and 3.1 Fast previews. Dates, the replacement id and what Sume lists.

The Veo 3.1 and Veo 3.1 Fast preview models in the Gemini API shut down on October 22, 2026, and Google's listed replacement is gemini-omni-1.1-flash. Google describes the listed dates as the earliest possible shutdown dates, so any request that still names veo-3.1-generate-preview or veo-3.1-fast-generate-preview may stop working from then, so there are about three weeks left from today. The Lite preview has the same date.
Sume's video catalog does not carry a Veo id. The Sume-side model that lines up with Google's replacement is gemini-omni-flash-1.1, which accepts text, first and last frame, reference media and a video edit input. This post lists the Google dates first, then the differences you will hit when a Veo-shaped request moves to Omni.
Which Veo ids end on October 22, 2026?
Google's deprecations page lists three Veo 3.1 previews with the same shutdown date and the same replacement. The older Veo 3.0 and 2.0 ids already had a June 30, 2026 shutdown, so they should be gone from any live code.
| Model id | Released | Shutdown | Replacement |
|---|---|---|---|
veo-3.1-generate-preview | October 15, 2025 | October 22, 2026 | gemini-omni-1.1-flash |
veo-3.1-fast-generate-preview | October 15, 2025 | October 22, 2026 | gemini-omni-1.1-flash |
veo-3.1-lite-generate-preview | March 31, 2026 | October 22, 2026 | gemini-omni-1.1-flash |
gemini-omni-1.1-flash | August 27, 2026 | No shutdown date announced | None |
What changes when you move from Veo to Omni?
Veo's own guide, read the same day, describes 4, 6 or 8 second clips, 1080p and 4K only at 8 seconds, 16:9 or 9:16, up to three reference images, and extension in 7 second steps. Omni Flash 1.1 on Sume is documented as 3 to 10 seconds at 360p, 720p, 1080p or 4K, 16:9 or 9:16, with native synced audio that cannot be turned off.
Two habits need checking. Extending a finished clip is a Veo feature, and Google's own changelog also lists video extension for Omni Flash, but Sume's docs describe no extend input: you request a new clip from a first frame, or use the video_url edit input on the Video Router. And a request that sets generate_audio: false is rejected for Omni rather than quietly producing silent video.
- Duration: any whole second from 3 to 10, not the 4/6/8 set.
- References: up to 10 images and up to 3 reference clips of at most 3 seconds each, addressed as
<IMAGE_REF_0>and<VIDEO_REF_0>in the prompt. - Edit: a
video_urlplus a prompt, withresolutionoptional (default 720p) and noaspect_ratioorduration.
What should you do before the date?
Pin the new id in a staging path now, run your five or six real prompts, and compare them by eye before the old id disappears. If you do not need to choose a family, sume/auto lets Sume pick the family, but Sume never discloses which family ran, so pin gemini-omni-flash-1.1 when you need a repeatable result.
Read GET /v1/videos/models for the live resolutions, durations and pricing_skus instead of copying numbers from a blog. Sume bills provider list times 1.25 and reserves that amount when the job is submitted.
Limits of this comparison
Google's page says nothing about quality differences between Veo 3.1 and Omni Flash, and this post does not either. It also covers only the Gemini API ids; Google's other platforms keep their own lifecycle pages. Check both docs.sume.com pages and Google's deprecations page again close to the date, because shutdown dates can move.
Sources
Related posts
More in Models
- Veo seed doesn't make output repeatable; Sume rejects seed
Google says Veo's seed only slightly improves determinism. Sume's v1 video models report seed false and reject the field; keep the output file to get a repeat.
- Which AI video model gives 1080p on Sume, and which stop at 768p?
Sume lists 1080p for Seedance, Wan 3.0, Kling 3.0 and Auto; MiniMax H3 is native 480p or 768p in the panel. Resolution table by model, with the API check.
- Voxtral TTS: open weights under CC BY-NC versus the paid API
Mistral's Voxtral TTS has open weights under CC BY-NC 4.0 and a paid API. What the licence means for a product, and what Sume offers instead.
- Wan 2.2 weights: Apache 2.0, 5-second 480p/720p; Sume hosts wan-3.0
Wan 2.2 T2V-A14B is Apache 2.0 with open weights and makes 5-second 480P or 720P clips. Sume does not host Wan 2.2; its Wan row is wan-3.0, 2 to 30 seconds.
Written by Sume