Veo 3.1 Lite vs Fast: half the price, minus three features
On Google’s Gemini API price list, a second of 720p video with audio costs $0.10 on Veo 3.1 Fast and $0.05 on Lite. Length and aspect ratios are the same. Google has scheduled the Gemini API previews of both models to shut down on October 22, 2026 and names Gemini Omni Flash as the replacement. We checked Google’s official docs on October 22, 2026.
Google’s Veo 3.1 video model comes in three price tiers: Veo 3.1, Fast and Lite. Lite costs half as much as Fast. In exchange it drops three things: 4K output, reference images and video extension.
Google has also set October 22, 2026 as the shutdown date for the Gemini API previews of all three models. We pulled only the price and feature rules from Google’s own pages, including the Gemini API pricing page, the Veo docs and the deprecations page. Each section links to its source.
Veo 3.1 Fast vs Lite on one sheet
Veo 3.1 Fast
- Per second with audio · 720p
- $0.10$0.80 for an 8-second clip
- 1080p · 4K
- $0.12 · $0.30
- Reference images · extension
- Up to 3 · 7 s steps
- Gemini API shutdown
- Oct 22, 2026
Veo 3.1 Lite
- Per second with audio · 720p
- $0.05$0.40 for an 8-second clip
- 1080p · 4K
- $0.08 · none
- Reference images · extension
- None · none
- Gemini API shutdown
- Oct 22, 2026
Gemini Omni Flash
- Per second with audio · 720p
- ~$0.10Google’s named replacement · token billing
- 1080p · 4K
- Upscaled · no price listed
- Reference images · extension
- Yes · 10 s steps to 40 s
- Gemini API shutdown
- Stable · none announced

Per second, Lite is half of Fast at 720p and two-thirds at 1080p
The Gemini API charges for Veo 3.1 per second of video with audio (Gemini Developer API pricing). None of the three is available on the free tier.
| Model | 720p | 1080p | 4K |
|---|---|---|---|
| Veo 3.1 | $0.40 | $0.40 | $0.60 |
| Veo 3.1 Fast | $0.10 | $0.12 | $0.30 |
| Veo 3.1 Lite | $0.05 | $0.08 | None |
| Gemini Omni Flash | ~$0.10 | — | — |
- An 8-second 720p clip costs $0.80 on Fast and $0.40 on Lite. Standard Veo 3.1 costs $3.20.
- 1080p costs $0.12 a second on Fast and $0.08 on Lite. Both models make 1080p only at 8 seconds.
- Only Fast outputs 4K, at $0.30 a second and only at 8 seconds.
When Google announced Lite on March 31, 2026, it said the model cost less than 50% of Veo 3.1 Fast at the same speed (Build with Veo 3.1 Lite). The same post said Fast prices would come down on April 7. On the October 22 price list, Lite is exactly half of Fast at 720p.
What Lite leaves out: 4K, reference images and extension
The Gemini API Veo docs list the request options for both models side by side (Generate videos with Veo 3.1 in Gemini API).
| Feature | Fast | Lite |
|---|---|---|
| 4K output | Yes | No |
| Reference images (up to 3) | Yes | No |
| Video extension (720p) | Yes | No |
| First and last frame | Yes | Yes |
| Length | 4, 6, 8 s | 4, 6, 8 s |
| Aspect ratio | 16:9, 9:16 | 16:9, 9:16 |
| Audio | Always on | Always on |
- Reference images show the model a person, character or product so it can preserve that appearance in the video. Fast takes up to three. Lite takes none.
- Extension adds 7 seconds at a time to a video you made with Veo, up to 20 times. It works only on Fast and only on 720p videos.
- Both models can start and end on frames you provide.
- Both make 4-, 6- or 8-second clips in 16:9 or 9:16 at 24 fps, and both always generate audio.
Google pitches them differently. The Veo docs say Fast suits backend services that generate ads programmatically, tools for rapid A/B testing of creative concepts and apps that need to produce social media content quickly. The Lite model page calls it the best price for developers and enterprises building high-volume video applications (Veo 3.1 Lite preview).
The Gemini API previews are set to shut down on October 22
Google’s Gemini API deprecations page lists October 22, 2026 as the shutdown date for the Veo 3.1, Veo 3.1 Fast and Veo 3.1 Lite previews (Gemini deprecations). The recommended replacement for all three is gemini-omni-1.1-flash. A notice on the Lite model page also points to the GA models on Google Cloud’s Gemini Enterprise Agent Platform.
Google describes these as the earliest announced shutdown dates. The page doesn’t say at what time the models go offline.
We queried the Gemini API’s model information at 9 a.m. UTC on October 22. All three previews still responded. Five Veo 3.0 and Veo 2.0 models whose shutdown dates have passed returned a model not found error.
Where to go next: Omni Flash or Google Cloud
Gemini Omni Flash became generally available on August 27, 2026 as a video generation and editing model (Gemini Omni Flash). Its model code is gemini-omni-1.1-flash. It makes 3- to 10-second videos with audio and lets you refine them in conversation. It outputs 360p, 720p, 1080p and 4K, with 1080p and 4K upscaled, in 9:16 or 16:9. It can work subjects from reference images into a video, and extension adds 10 seconds at a time up to 40 seconds (Generate and edit videos with Gemini Omni Flash).
Omni Flash bills by token. Google counts a second of 720p video as 5,792 tokens and puts the effective price at about $0.10 a second. It also uses a different API from Veo, the Interactions API.
On Google Cloud, the Gemini Enterprise Agent Platform docs list veo-3.1-fast-generate-001 as generally available since November 17, 2025 and veo-3.1-lite-generate-001 as a preview released on April 2, 2026 (Veo 3.1 | Gemini Enterprise Agent Platform). Both are listed only in the us-central1 region, with English as the prompt language.
Google Cloud’s prices with audio match the Gemini API figures (Agent Platform Pricing). It also lists cheaper video-only prices: $0.08 a second for Fast at 720p and $0.03 for Lite.
Which one to pick
The rest of this section is our analysis, not Google’s documentation.
For an 8-second vertical clip at 720p, Lite costs $0.40 and Fast $0.80. Omni Flash comes to about $0.80 by Google’s own estimate. If you used Fast, the price barely changes. If you used Lite, it doubles.
To keep Lite pricing you’d have to move to Lite on Google Cloud, which is still a preview there. If you need the same person or product across several shots, pick a model that takes reference images: Fast or Omni Flash. To carry a scene past 8 seconds you need extension, which Fast and Omni Flash offer.
Google hasn’t evaluated prompts in other languages for either model. The Veo docs and the Omni Flash guide both say only English is fully supported. If your clips have dialogue in another language, test a short one first.
What we couldn’t confirm
- We didn’t generate any videos for this article, so we didn’t measure differences in quality, speed or non-English dialogue. Every figure comes from Google’s documentation.
- The docs don’t say what time on October 22 the previews shut down. An image model on the same deprecations page, gemini-3.1-flash-image-preview, carries a June 25 shutdown date and still responded on October 22.
- The Gemini API’s model information reports an input limit of 480 tokens for the three previews. The docs’ model table says 1,024 tokens. We didn’t test which one applies.
- The price list gives no 1080p or 4K figure for Omni Flash. Google’s per-second estimate covers 720p only.
- Google Cloud’s pages don’t agree with each other. The Fast model table lists no 4K output, but the price list has a Fast 4K price. The Lite model table says extension is supported, while the Gemini API docs say Lite has none. The price list gives its unit as 1 count.
- This article was written by Sera Studio. Sera Studio’s Motion Studio offered Veo 3.1 Fast and Lite through the Gemini API and removed them from its model list on November 13, 2026, after the previews shut down.


