# Google Flow: Veo 3.1 Lite vs Omni Flash 1.2, Which to Pick

> Veo 3.1 Lite or Omni Flash 1.2 in Google Flow? Credit costs per clip and per second, what each can edit, and which models retire December 5, 2026.

_Source: https://professionalstoolkit.com/articles/google-flow-veo-vs-omni — The Professional's Toolkit · updated November 2026_

---

As of November 2026, Google Flow offers two video models: Omni Flash 1.2 and Veo 3.1 Lite. The choice comes down to how each one bills and what you plan to do with the clip afterward. Veo 3.1 Lite charges a flat 10 credits per clip (5 on Ultra) for a 4, 6 or 8-second 720p shot, and Flow cannot edit or extend what it makes. Omni Flash 1.2 bills by the second, 1 credit at 360p and 3 at 720p, and it is the only model that edits, extends and runs to 15 seconds on paid plans.[1] Our rule of thumb is short. Lite for finished one-off shots; Omni for anything you will rework.


**30 credits** — One 10-second 720p Omni Flash 1.2 clip in Flow by our arithmetic, against 10 credits for a Veo 3.1 Lite clip (5 on Ultra).



- **10 credits** — Veo 3.1 Lite per clip in Flow (5 on Ultra)
- **3 credits/s** — Omni Flash 1.2 at 720p, seconds 1-10
- **Dec 5, 2026** — Veo 3.1 Fast, Quality and Omni Flash 1.1 retire in Flow


## What changed in Flow's model menu this fall

Flow's model list has been shrinking. **Veo Extend was retired on November 5, 2026**, so extending a clip now needs a paid plan and Omni Flash 1.2. Google also retires Veo 3.1 Fast, Veo 3.1 Quality and Omni Flash 1.1 in Flow on December 5, 2026; until then they stay available at standard credit rates, and Google recommends moving to Omni Flash 1.2 or Veo 3.1 Lite. If a saved workflow depends on one of the leaving models, rebuild it now rather than on the morning it breaks.

The older models still carry their own prices for a few weeks. Until December 5, 2026, Veo 3.1 Fast takes 20 credits per generation (10 on Ultra) for a 4, 6 or 8-second clip. Until December 5, 2026, Veo 3.1 Quality takes 100 credits for an 8-second clip. Until December 5, 2026, Omni Flash 1.1 charges 4 to 15 credits per clip depending on length and resolution, and editing with it costs a flat 40 credits. None of those rates survive the retirement date. The two models that remain are the real decision.


> ⚠ **Older models leave December 5:** Google retires Veo 3.1 Fast, Veo 3.1 Quality and Omni Flash 1.1 in Flow on December 5, 2026. Veo Extend was already retired on November 5, 2026; extending now needs a paid plan and Omni Flash 1.2.


## How the two models charge you

Both remaining models are open on every plan, including a free account with its 50 Flow credits a day; plans change the clip length ceiling, the Ultra discount on Lite and access to extension. The table lists Flow Help's credit costs for each option as of November 2026, with plan gates noted in the row where they apply.

| Option | Clip length | Resolution | Flow credits |
|---|---|---|---|
| Veo 3.1 Lite, standard | 4, 6 or 8 s | 720p | 10 per clip (5 on Ultra) |
| Veo 3.1 Lite, Lower Priority mode | 4, 6 or 8 s | 720p | 0 (Ultra 20x and family plan managers only) |
| Omni Flash 1.2, seconds 1-10 | 3 to 10 s | 360p / 720p | 1 / 3 per second |
| Omni Flash 1.2, seconds 11-15 (paid plans) | up to 15 s | 360p / 720p | 2 / 6 per second |
| Omni Extend (paid plans) | adds 3 to 10 s | 360p / 720p | 1 / 3 per added second |

The difference is the billing shape. A Lite clip costs the same whether it runs 4 or 8 seconds, while Omni grows with every second and doubles its rate past the tenth. By our arithmetic, one 10-second 720p Omni clip costs 30 credits and a 15-second 720p clip costs 60, against 10 credits for any Lite clip. At 360p, a 10-second Omni clip costs 10 credits and a 15-second one costs 20, which puts Omni drafts on the same footing as Lite.[2]

Credits are charged per generation, not per request, so a request that returns several clips spends several times over. Flow may also rate-limit generations on any plan, especially after many in one day. Unused daily and monthly credits expire at each refresh, whichever model spent them, so a model that saves credits only helps if you use the savings inside the same cycle. Our [credits guide](/articles/google-flow-credits) covers the remaining rules.

## When Veo 3.1 Lite is the right call

Flow Help describes Veo 3.1 Lite as a dependable, low-cost baseline for standard 720p creation. That fits a shot you can describe completely in one prompt and will not touch again: a product turning on a table, a short establishing scene, a transition between two stills. Lite handles Text to Video and Frames to Video with a first frame, or a first and last frame, at 4, 6 or 8 seconds in both aspect ratios. Ingredients to Video works too, but only at 8 seconds. Veo generates sound effects, ambient noise and dialogue natively, though DeepMind says natural speech in short segments is still an area of active development.

The cost case is plain. By our arithmetic, a free account's 50 daily credits cover five Lite clips a day, a Google AI Pro month's 1,000 credits cover 100 clips, and Ultra 5x's 10,000 credits cover 2,000 clips at the Ultra rate of 5 credits. Ultra 20x subscribers can drop the price to zero with Lower Priority mode. Quality is not the trade you might expect, either. In Google's own evaluation in its April 2026 model card, Veo 3.1 Lite won 54.6% of text-to-video and 47.2% of image-to-video comparisons against Veo 3.1 Fast, which itself retires in Flow on December 5, 2026.

The limits are hard ones. Lite cannot edit video, cannot extend a clip, and stops at 8 seconds and 720p. A shot that misses means a fresh generation at 10 credits, well under the 30 credits of a single 10-second 720p Omni clip. That makes trial and error cheap on Lite, as long as each attempt is a full re-roll of the same short shot rather than a fix to one detail inside it.


> 💡 **Draft Omni clips at 360p:** Omni Flash 1.2 at 360p costs a third of 720p. On Plus, Pro or Ultra, upscaling a clip of 10 seconds or less to 720p or 1080p costs 0 credits.


## When Omni Flash 1.2 earns its per-second price

Omni Flash 1.2 is the model for clips that will change after the first take. Flow Help positions it for higher visual quality, longer videos and duration-based pricing, and it is the only Flow model that does video-to-video editing: insert, remove and camera changes, with up to three conversational turns of context. Paid plans add clips up to 15 seconds and Omni Extend, which adds 3 to 10 seconds per step, guided by up to three image references.

Resolution is the lever that keeps Omni affordable. Flow lets everyone draft at 360p for a third of the 720p cost, and Plus, Pro and Ultra subscribers can then upscale a chosen clip to 720p or 1080p for 0 credits when it runs 10 seconds or less, plus 1 credit per extra second. Upscaling to 4K is for Ultra only, at 50 credits for a 10-second clip by our arithmetic. Prompting differs too. DeepMind's prompt guide says Veo needs precise instructions while Omni needs less prescriptive prompts; our [prompt guide](/articles/google-flow-prompt-guide) covers both styles.

Omni also carries people across clips. A character bundles one or two images and a voice, called into a prompt with its name, and voice references work only on Omni generations that use ingredients. Two gates apply before you plan around any of this. Frames to Video with Omni and extension work on the web only, not in the mobile app. Editing uploaded videos is unavailable in the EEA, the UK and Switzerland.


> 💡 **Use Lite for fixed shots:** Veo 3.1 Lite costs 10 credits per clip (5 on Ultra) whether it runs 4 or 8 seconds, so choose 8 seconds when the shot allows it. It cannot be edited or extended in Flow afterward.


## Picking a model shot by shot

Most projects mix both models, so decide per shot rather than per project. Flow's Scenebuilder arranges, reorders and trims clips into one scene, so a Lite establishing shot and an Omni dialogue shot with an edited background can sit in the same sequence and download together. Four checks settle almost every case:

- **Will you edit or extend it?** Use Omni Flash 1.2. Lite does neither.
- **Is it longer than 8 seconds?** Omni again; 11 to 15 seconds needs a paid plan.
- **Is it a fixed 720p shot of 8 seconds or less?** Lite, at 10 credits (5 on Ultra).
- **Does a recurring character need the same face and voice?** Omni, with a saved character.

Volume changes the math more than taste does. By our arithmetic, a Google AI Pro month's 1,000 credits buy 100 Lite clips, or 33 ten-second 720p Omni clips, or 16 fifteen-second 720p Omni clips. A team that needs dozens of finished shots a month leans on Lite. One that edits a handful of hero clips leans on Omni and drafts at 360p first. Free users have a further constraint: video generation may be unavailable from roughly 2 PM to 5 PM UTC, whichever model you choose.

## Veo and Omni outside Flow

The same names live elsewhere at different terms, and none of these figures are Flow prices. The Gemini app replaced Veo with Gemini Omni, which needs a Google AI plan and runs 3 to 10 seconds, or 15 on Ultra. In the Gemini API, Veo 3.1 Preview was shut down on October 22, 2026, and developers are pointed to gemini-omni-1.2-flash; when the Omni Flash preview launched in June 2026, Google priced it at $0.10 per second of output video.[3] Businesses can still reach Veo 3.1 on Google Cloud, where the pricing page lists $0.40 per second with audio at 720p or 1080p, and Veo 3.1 Lite (a Preview model there) at $0.05 per second with audio at 720p.[4] That Veo 3.1 runs in one US region, on provisioned throughput or fixed quota rather than pay-as-you-go, with retirement listed as May 2027 or later.


## FAQ

**Does the model I pick change the watermark?**

No. Every Flow output, made with Veo or Omni, carries an invisible SynthID watermark, and Omni content also carries C2PA Content Credentials. A visible watermark is an optional toggle, applied automatically in India, South Korea and Vietnam unless you are on Google AI Ultra 20x.

**Can I use Omni Flash 1.2 in the Flow mobile app?**

Text to Video works in the apps, but Frames to Video with Omni Flash 1.2 and clip extension are web only, as of November 2026. Google recommends a desktop Chromium browser for the fullest editing.

**Do failed generations cost credits?**

Flow returns credits for failed or safety-blocked generations within one hour. If Veo produces poor audio, Flow does not return the video and refunds the credits, unless the Return silent videos setting is on; then you get a silent clip and are charged.


## The bottom line

In November 2026 the Flow decision is a two-way choice, and it is mostly about billing shape. Veo 3.1 Lite gives you a fixed-price 720p shot of up to 8 seconds that you accept as generated, which makes it the cheaper route for volume. Omni Flash 1.2 costs more per finished second but is the only path to editing, extension, recurring characters and 15-second clips. Draft Omni at 360p, upscale the keepers on a paid plan, and move any workflow built on Veo 3.1 Fast, Quality or Omni Flash 1.1 before they retire on December 5, 2026.



## Sources

[1] Learn about Google Flow models & supported features - Google Flow Help — https://support.google.com/flow/answer/16352836?hl=en (2026-11)
[2] Manage your Google Flow credits - Computer - Google Flow Help — https://support.google.com/flow/answer/16526234?hl=en (2026-11)
[3] Google blog: Gemini Omni Flash and Nano Banana 2 Lite for developers — https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-omni-flash-nano-banana-2-lite/ (2026-11)
[4] Google Cloud: generative AI pricing — https://cloud.google.com/vertex-ai/generative-ai/pricing (2026-11)
[5] Google AI plans (Google One) — https://one.google.com/intl/en_us/about/google-ai-plans/
[6] Google Terms of Service — https://policies.google.com/terms?hl=en