Gemini Omni Flash for client work: app, Flow, Shorts, or an API?

Gemini Omni Flash is in five places. Who each is for, what is free, and when a client project needs an API with job ids rather than an app.

5 min readSume
All posts

For client work that needs repeatable batches, a record of what was run, and files you can hand over, use an API; for a quick personal clip, use the Gemini app, Flow, or YouTube Shorts and Create. Google offers Gemini Omni Flash in all four of those apps plus the Gemini API, and Sume exposes it as gemini-omni-flash-1.1.

Facts below are from Google's I/O 2026 roundup, the YouTube Blog, the Gemini API pricing and release notes, read 2026-10-02.

Where can I use Gemini Omni Flash?

Google's pages place it in the Gemini app and Flow for Google AI Plus, Pro and Ultra subscribers, and in YouTube Shorts Remix and the YouTube Create app at no cost for users 18 and older. The release notes say gemini-omni-1.1-flash became generally available on August 27, 2026, with extension, interpolation between images, and 720p default with 1080p and 4K output; the older preview endpoint was set to be deprecated on September 30, 2026.

Where Gemini Omni Flash lives, read 2026-10-02
SurfaceWho gets itAPI or files?
Gemini appGoogle AI Plus, Pro, UltraApp only
FlowGoogle AI Plus, Pro, Ultra; free daily credits existApp only
YouTube Shorts Remix and CreateUsers 18+, no costInside YouTube
Gemini APIBilled per useYes
Sume Video RouterSume balanceYes, async jobs

What does the API cost?

Google's pricing page lists Omni Flash input at $1.50 per 1M tokens and video output at $17.50 per 1M tokens, with billing at 5,792 tokens per second of 720p video, about $0.10 per second. On Sume the docs say billing is provider list times 1.25 per output second by resolution, so budget a quarter more than Google's list.

When does a client project need an API?

Reach for one when any of these is true: you need many variants (a hook test across a product line), you need to prove later which prompt made which file, you need a webhook when the work is done, or several people share the job. An app gives you a result in a tab. An API gives you a job id, a status, and a hosted URL.

Sume jobs are readable by the member whose key created them, and the Jobs and results page tells integrations to store job ids so work survives restarts.

What should I check before choosing Sume?

Sume's envelope for this model is 3-10 seconds at 360p to 4K in 16:9 or 9:16, with audio always on and no audio reference input. Edit mode takes a video_url and cannot be combined with reference images. If you need the YouTube-side Remix features, watermarks inside YouTube's metadata, or Flow's own tools, those stay inside Google's apps.

A minimal call is in the Video Router docs; the rest of the contract is in Video generation.

curl -X POST https://api.sume.com/v1/videos \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: client-a-hook-01" \
  -d '{
    "model": "gemini-omni-flash-1.1",
    "prompt": "A courier hands a parcel to a smiling customer at a door, 9:16 handheld",
    "aspect_ratio": "9:16",
    "duration": 6,
    "resolution": "720p"
  }'

Sources

Related posts

More in Comparisons

All Comparisons posts

Written by Sume