Correction, 2026-10-02: this page told you to use Omni Flash for drafts and full Veo 3.1 for the shot that matters, which assumed a quality difference we have not measured; it now compares facts and suggests running the same prompt on both. It said Omni "wins on the loop" at 51 credits per 3-second draft, but Veo 3.1 Lite is cheaper per draft (27 credits for 4 seconds with sound, 17 silent), and its draft workflow compared Omni only against Veo Standard. The FAQ called Lite's 720p "the trade", but Omni Flash also returns 720p on 8frame. It said Veo "renders natively at higher resolution" (Google's pricing page lists Veo's resolutions, not how they are made) and implied that seven reference images keep a character consistent on Omni, which we have not tested. The closing line put Veo on 8frame from 27 credits; the cheapest Veo run is 17.
Correction, 2026-10-01: we called Veo 3.1 Standard one of the most expensive defaults on the canvas. It is not the default: a new Veo node opens on Veo 3.1 Fast (168 credits for 8 seconds with audio), and Standard is a switch in the picker. We also left out Veo's reference option (1 to 3 images) when comparing inputs, and now say what each model returns on 8frame (Omni Flash 720p, Veo 3.1 1080p, no 4K).
Correction, 2026-09-30: we quoted our own production volumes (which we do not publish), placed Omni Flash between Veo Fast and Standard on price (it costs less than Fast), said four Standard clips use half a Starter month (they use nearly two), and made quality claims we have not tested. All fixed.
Gemini Omni Flash and Veo 3.1 are both Google video models with sound, and both run on the 8frame canvas. We have not compared their output quality, so this page sticks to what is fixed. On 8frame an 8-second Omni clip is 135 credits at 720p, with sound always on. Veo 3.1 is 54 credits on Lite (720p), 168 on Fast (1080p, the tool's default) and 448 on Standard (1080p), with sound optional. Omni runs 3 to 10 seconds and takes up to seven reference images or a clip to edit; Veo runs 4, 6 or 8 seconds and takes a start image with an optional end frame, or up to three reference images. On Google's own API, Omni Flash is roughly $0.10 per second at 720p, the same as Veo 3.1 Fast at 720p, against $0.40 for Veo 3.1 Standard. To choose for your shot, run the same prompt on both.
TL;DR
- Omni Flash on 8frame: 135 credits for 8 seconds, 3 to 10 seconds per clip, 720p, sound always on, 16:9 or 9:16. Inputs: a prompt, a start image, 1 to 7 reference images, or a 3- to 10-second clip to edit.
- Veo 3.1 on 8frame: 4, 6 or 8 seconds (always 8 with a start image), 16:9 or 9:16, sound optional. The tool opens on Fast (168 credits for 8 seconds with sound, 1080p); Lite is 54 (720p), Standard 448 (1080p).
- Cheapest per clip: Veo 3.1 Lite, at the same 720p as Omni. Longest clip: Omni, at 10 seconds. 1080p: Veo Fast or Standard only.
- Quality: not measured by us. A same-prompt test at 4 seconds on Omni, Veo Lite and Veo Fast costs 179 credits.
The price table
Every number below is what the 8frame canvas charges, in credits, with sound on unless stated. A credit is about $0.01 at pack rate.
| Model | 3s | 4s | 5s | 6s | 8s | 10s | Silent 8s | Output |
|---|---|---|---|---|---|---|---|---|
| Gemini Omni Flash | 51 | 68 | 85 | 102 | 135 | 169 | Not available | 720p |
| Veo 3.1 Lite | n/a | 27 | n/a | 41 | 54 | n/a | 33 | 720p |
| Veo 3.1 Fast (Veo default) | n/a | 84 | n/a | 126 | 168 | n/a | 112 | 1080p |
| Veo 3.1 Standard | n/a | 224 | n/a | 336 | 448 | n/a | 224 | 1080p |
Omni runs every whole second from 3 to 10 at about 16.9 credits a second. The cheapest Veo run is 17 credits: 4 silent seconds on Lite.
Three things fall out of that table. First, at 8 seconds Omni Flash costs less than Veo Fast and more than Veo Lite, and Lite returns the same 720p. Second, Omni is the one that goes past 8 seconds (to 10) or below 4 (to 3). Third, Veo's silent option is a real lever: if the clip is getting a music bed in the edit anyway, turning Veo's sound off halves the bill on Standard (448 to 224), cuts Fast by a third (168 to 112) and Lite by about 40% (54 to 33). Omni has no silent option on 8frame.
How they differ, fact by fact
| On 8frame | Gemini Omni Flash | Veo 3.1 |
|---|---|---|
| Length | 3 to 10 s, any whole second | 4, 6 or 8 s; always 8 s with a start image |
| Shape | 16:9 or 9:16 | 16:9 or 9:16 |
| Output | 720p | 720p on Lite, 1080p on Fast and Standard |
| Sound | Always on | On by default, can be turned off |
| Start image | Yes | Yes |
| End frame | No | Yes, at 8 s |
| Reference images | 1 to 7, not combined with a start image | 1 to 3 on the reference option, 8 s |
| Edit an existing clip | Yes, 3 to 10 s, keeps its length | No |
| Median generation time | About 35 s for 3- to 5-second clips; about 49 s for an edit | No measured figure yet |
If the deliverable has to be 1080p, Veo Fast or Standard is the route on 8frame: Omni Flash and Veo Lite both return 720p, and no video model on 8frame outputs 4K. Google's API goes further (1080p and 4K upscales for Omni; 4K for Veo Standard at $0.60 a second and Fast at $0.30), but those are API features, not 8frame ones.
If the shot runs longer than 8 seconds, Omni makes up to 10 in one clip. Past that, it is an edit on either model.
If you have more than three reference images, only Omni takes them all, up to seven. If you need a fixed first and last frame, only Veo does that on 8frame.
If you have footage to change, Omni's video edit takes a 3- to 10-second clip and a prompt. Veo has no edit mode on 8frame.
If the sound will be replaced in the edit, Veo's silent option saves credits; Omni charges for sound every time.
The catches worth knowing before you commit
Omni bills sound whether you want it or not. Over a batch of 40 eight-second clips, that is 5,400 credits on Omni against 4,480 on silent Veo Fast or 1,320 on silent Veo Lite.
Omni caps at 10 seconds per generation on 8frame. Google's API extends Omni clips in 10-second steps to 40 seconds in total, a 1.1 feature announced on 2026-08-27, but the 8frame tool makes 3- to 10-second clips and does not extend them.
Veo locks to 8 seconds with a start image. The 4- and 6-second lengths are for text prompts only.
Uploaded-video edits have a regional limit on Google's API. Per Google's docs, editing or extending an uploaded video on its API is not available in the European Economic Area, Switzerland or the UK; editing a video the model generated is.
Veo Standard is the expensive tier. At 448 credits for 8 seconds (56 a second), two Standard clips take 896 of a Starter month's 1,000. A new Veo node opens on Fast (168); switch to Standard when you have a reason to.
Run the same prompt on both
We have not compared Omni Flash and Veo 3.1 on output quality, and a ranking would not tell you how they handle your shot anyway. Put both nodes on one board, give them the same prompt and the same shape, and compare:
- Text prompt, 4 seconds: Omni 68, Veo Lite 27, Veo Fast 84, Veo Standard 224. All four: 403 credits; without Standard, 179.
- Start image, 8 seconds (Veo's only length with an image): Omni 135, Veo Lite 54, Veo Fast 168, Veo Standard 448. All four: 805 credits; without Standard, 357.
Either test fits inside the 1,000 credits of the $19 Starter plan. Google's Omni guide says Omni Flash tries to create "a video with a few different shots" by default, so if you want one continuous shot, say so in the prompt you give both models. Watch the results where they will be shown, with the sound on, since sound is part of what you are comparing. Each rerun is a new generation, so expect a different take every time, and failed generations are refunded automatically.
FAQ
Is Gemini Omni the same thing as Veo? No. They are separate Google model lines. Omni is the multimodal model that generates picture and sound together and can edit video; Veo is the dedicated video family with Standard, Fast and Lite tiers. In the Gemini app, Google says "Gemini Omni will replace Veo"; on Google's API and on 8frame, both are available.
Which one is cheaper? On 8frame, for 8 seconds with sound: Veo 3.1 Lite at 54 credits, then Omni Flash at 135, then Veo Fast at 168, then Veo Standard at 448. Lite and Omni both return 720p; Fast and Standard return 1080p. The cheapest Veo run is 17 credits (4 silent seconds on Lite), the cheapest Omni run 51 (3 seconds).
Which one looks better? We have not measured it. Run the same prompt on both: the 4-second test above costs 179 credits without Veo Standard.
Can I run both on the same prompt? Yes. Both are nodes on the same canvas, so you can give them one prompt and compare the results side by side. Best AI video generator 2026 covers the other video models on the canvas.
Sources
- Omni model spec (16:9 default and 9:16; 360p and 720p, 1080p and 4K upscaled; extension in 10-second steps to 40 seconds; 10-second upload cap; regional note; "a few different shots" prompt note): ai.google.dev/gemini-api/docs/omni, page last updated 2026-09-23, checked 2026-10-02
- API prices (Omni Flash $17.50 per 1M output video tokens, about $0.10 per second at 720p; Veo 3.1 Standard $0.40 at 720p and 1080p, $0.60 at 4K; Fast $0.10, $0.12 and $0.30; Lite $0.05 and $0.08): ai.google.dev/gemini-api/docs/pricing, page last updated 2026-10-01, checked 2026-10-02
- Omni Flash priced the same as Veo 3.1 Fast on the API: Google's public preview post, 2026-06-30, checked 2026-10-02
- Omni 1.1 announcement: blog.google, 2026-08-27, checked 2026-10-02
- Gemini app: Gemini video generation ("Gemini Omni will replace Veo in the Gemini app"), checked 2026-10-02
- 8frame prices: backend tool registry (calculateGeminiOmniFlashCost, calculateVeo31Cost), read 2026-10-02; lengths, inputs, reference limits and defaults: front-end tool registry, read 2026-10-02; Veo shapes and output resolution: 8frame Veo 3.1 integration, checked 2026-10-01; Omni's measured 720p output and generation times: 8frame production data, read-only, 2026-09-30
Both models sit on the same canvas, so the comparison that settles it is your own prompt on both. Run Gemini Omni Flash from 51 credits and Veo 3.1 from 17 on 8frame, from $19/month.
Try it without signing up: the 8frame canvas opens on a real board. The canvas is free and unlimited; generation is paid from $19/month.