Gemini Omni Flash Preview

Retired

Preview endpoint of Gemini Omni Flash, Google's conversational video generation and editing model; the model page lists it under Versions as 'Preview: gemini-omni-flash-preview' alongside the stable gemini-omni-1.1-flash. Google's deprecations table, read 2026-10-07, gives release June 30, 2026, shutdown October 22, 2026 and gemini-omni-1.1-flash as the recommended replacement; until 2026-10-23 the row was held at deprecated, because on every read up to 2026-10-22 it was NOT gray, and gray is Google's own marker for an endpoint that is already off. Until 2026-10-06 retires_on carried September 30, 2026, the date the same table gave on the 2026-09-01 read. Last published rates, identical to the stable endpoint: input $1.50 per 1M labelled text / image / video / audio, and two output rates — text and thinking tokens $9.00/1M, video tokens $17.50/1M — which price_output_per_mtok held (the video rate) until the pricing section disappeared. Google states billing is on total output token consumption at 5,792 tokens per second of 720p video, an effective ~$0.10 per second under Standard pricing. Paid tier only; no free tier and no context caching rate published. Context window 1,048,576 tokens; the output is video, so no max output token count exists. Added 2026-09-01 together with its replacement target, because a pointer to a row the catalog does not carry hard-fails check-refs (LEARNINGS #61b). Between the 2026-10-02 and 2026-10-06 reads the row was absent in Google's deprecations table at all (the Gemini Omni Flash section now lists only gemini-omni-1.1-flash, 'No shutdown date announced'), while the pricing page kept a 'Gemini Omni Flash Preview' section with the rates above and the model page still listed 'Preview: gemini-omni-flash-preview'; the fields were held as last published, and the 2026-10-07 read restored the row with the new October 22 date. CONFIRMED RETIRED 2026-10-23 on the only signal this catalog accepts for a Google shutdown: the gemini-omni-flash-preview row on the deprecations page now carries a literal class="row-gray" on its <tr> (page: Last updated 2026-10-22 UTC), and that page's own legend reads "Already-shutdown models are indicated with gray backgrounds"; check-lifecycle-drift.mjs reported Google stating "shut-down" against our "deprecated". Corroborated on the same read by the Gemini Omni Flash model page, whose Versions block now reads "Shut down: gemini-omni-flash-preview" (until 2026-10-22 it read "Preview: gemini-omni-flash-preview"). The pricing page no longer carries a Gemini Omni Flash Preview section or its October 22 deprecation warning, so the price fields are null rather than carried forward.

Gemini Omni Flash Preview by Google is tracked live in our LLM catalog, with a 1.0M (1.048.576-token) context window. It has been retired (shut down 22 Oct 2026); the recommended migration is Gemini Omni Flash.

Last verified: 24 Oct 2026 · sourced from official provider documentation

Retired on 22 Oct 2026

Provider
Google
Status
Retired
Input price
—
Output price
—
Cached input
—
Blended price
—
Context window
1.048.576 tokens (1.0M)
Max output
—
Modality
text, image, video, video-out
Open weights
No — API only
Knowledge cutoff
—
Released
30 Jun 2026
API string
gemini-omni-flash-preview

Source: Google official documentation ↗

FAQ

Gemini Omni Flash Preview — questions & answers

What is the context window of Gemini Omni Flash Preview?

Gemini Omni Flash Preview has a 1.048.576-token context window (1.0M).

Is Gemini Omni Flash Preview still available?

Gemini Omni Flash Preview has been retired and is no longer available. Its scheduled shutdown date is 22 Oct 2026. The recommended migration is Gemini Omni Flash.

Does Gemini Omni Flash Preview support image or vision input?

Yes — Gemini Omni Flash Preview accepts image input. Listed modalities: text, image, video, video-out.

Can I self-host Gemini Omni Flash Preview?

No — Gemini Omni Flash Preview is proprietary. Its weights are not publicly released, so it is available only through Google's API (or hosted partners), not for self-hosting.

What is the API model string for Gemini Omni Flash Preview?

The API model identifier for Gemini Omni Flash Preview is "gemini-omni-flash-preview" when calling Google's API.