Test an Omni edit prompt at 360p first: 23 cents vs 113 at 1080p

Gemini Omni Flash 1.1 edit mode lets you set resolution, so the cheap way to find a working edit prompt is to run it at 360p and only then re-run the winner at 1080p. For a 6-second output that is 23 cents per try at 360p against 113 cents at 1080p, a 5x difference.

What edit mode takes

Per the Video Router docs, video_to_video sends video_url and a prompt with the instructions. resolution is optional and defaults to 720p. There is no aspect_ratio and no duration in an edit, and you cannot combine video_url with image_url, end_image_url or the reference fields. The billing rule is the provider list times 1.25 per output second.

The docs do not state the edit output length, so the prices here assume a 6 s output. Confirm with usage.cost on the first poll.

curl -X POST https://api.sume.com/v1/video-router/generate \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: edit-test-360-001" \
  -d '{
    "model": "gemini-omni-flash-1.1",
    "prompt": "Replace the bottle with an apple. Keep everything else the same.",
    "video_url": "https://example.com/clip.mp4",
    "resolution": "360p",
    "mode": "async"
  }'

The price of a test round

Say you want to try six prompt wordings. At 360p that is six runs of 23 cents, or $1.38. The same six runs at 1080p would be $6.78. Then you render the chosen edit once at 1080p for $1.13, so the whole job costs $2.51 instead of $6.78 plus the final.

Sume rounds each clip up to the next whole cent, so a clip total is ceil(seconds x rate), not a sum of fractional cents.

What a 360p test can and cannot tell you

A low-resolution render shows whether the model found the object, replaced it and left the rest alone. It does not show fine texture or text. If your edit depends on small lettering, test at 720p instead; that is 75 cents for 6 s.

Use a different Idempotency-Key for each wording. The same key replays the original job, which is what you want for a network retry and not for a new prompt.

The same length at every tier

For reference, a 6-second Omni Flash clip at each resolution. Every price is the seconds times the billable rate, rounded up to a whole cent, as of 2026-10-08.

Limits to remember

These apply to every request on this page, from the Video Router and Video generation docs:

• Length is 3 to 10 whole seconds in generation modes; an edit takes no duration.
• Aspect ratio is 16:9 or 9:16; an edit takes no aspect ratio.
• Native synced audio is always on, and generate_audio: false is rejected.
• There is no bitrate_mode, no reference audio and no seed.
• Billing is the provider list times 1.25 per output second, reserved at submit and shown in usage.cost.