Veo 3.1
Veo 3.1 is a video generation model released by Google DeepMind on October 15, 2025, as an incremental update to Veo 3. It keeps the headline capability that defined its predecessor, native generation of synchronized audio alongside video. Google said the update brought "richer audio, more narrative control, and enhanced realism that captures true-to-life textures," with "stronger prompt adherence and improved audiovisual quality when turning images into videos" [2]. The update was distributed simultaneously across Google's consumer and developer surfaces: the Flow filmmaking tool, the Gemini app, the Gemini API, and Vertex AI [1][2][3]. A faster, cheaper variant, Veo 3.1 Fast, launched the same day, and Google added a lower-cost Veo 3.1 Lite in March 2026 [3][11]. As with every model in the family, Veo 3.1 outputs carry a SynthID watermark [3].
Google updated the model on January 13, 2026, adding native vertical output to its Ingredients to Video mode, improved 1080p output and a new 4K option [7][8]. In May 2026 Google announced Gemini Omni, which it said would replace Veo in the Gemini app, and the Veo 3.1 preview models in the Gemini API were shut down on October 22, 2026 [12][16][17]. As of October 30, 2026, Veo 3.1 models were still offered in Flow and as generally available models on Google Cloud's Gemini Enterprise Agent Platform, the successor to Vertex AI [13][14][18].
Background (Veo 3)
Google's video work runs through three earlier releases. The original Veo arrived at Google I/O in May 2024, followed by Veo 2 in December 2024, which added 4K output and an improved understanding of physics. Veo 3, announced at I/O in May 2025, was the step change: it generated audio natively, including dialogue, sound effects, and ambient noise tied to the visuals. Google DeepMind chief executive Demis Hassabis framed the moment by saying the field was "emerging from the silent era of video generation" [4][5].
Veo 3 became the engine behind Flow, Google's web-based tool for stitching AI clips into longer sequences. By the time Veo 3.1 shipped, Google said more than 275 million videos had been generated in Flow since the tool launched in May 2025, a count that included Veo 2 and Veo 3 generations [1][2]. Google presented Veo 3.1 as a model that "builds on Veo 3" rather than as a new generation, and it replaced Veo 3 as the model behind Flow [2][23].
New features
The most consistent theme across Google's announcements is that audio was extended to capabilities that previously produced silent output. At launch Google said it was, "for the first time," bringing audio to "Ingredients to Video," "Frames to Video" and "Extend" [2]. TechCrunch noted that Veo 3 already offered reference images, first-and-last-frame generation, and clip extension, and that Veo 3.1 added audio to all of them [1]. Alongside that, the company described "stronger prompt adherence and improved audiovisual quality when turning images into videos" [2]. In its developer announcement Google said the models "now generate richer native audio, from natural conversations to synchronized sound effects, and offer greater narrative control with an improved understanding of cinematic styles" [3].
The features break down as follows:
| Feature | What it does |
|---|---|
| Ingredients to Video | Uses multiple reference images to control characters, objects, and style; the Gemini API accepted up to three reference images of a character, object, or scene [2][3] |
| Frames to Video (first and last frame) | Generates a transition between a supplied start image and end image, now with audio [2][3] |
| Extend (scene extension) | Continues a clip into a longer sequence, generating each new segment from the final second of the previous one, allowing videos of a minute or more [2][3] |
| Insert | Adds an object or element into an existing scene in Flow, with the model handling shadows and lighting so it blends in [1][2] |
| Remove | Erases unwanted objects and reconstructs the background behind them; announced as coming after launch [1][2] |
Google said the audio additions were "experimental and actively improving" [2]. The object-removal tool was the main Flow feature not available on day one [1][2]. On November 25, 2025, Google said users could tap a pencil icon on any Flow clip to insert or remove elements, but that "object removal is rolling out next month and will be experimental" [20]. Insert and Remove were Flow features: Google said "Add object" and "Remove object" were not available in the Gemini API at launch [2].
Launch specifications
The Gemini API documentation published at launch described Veo 3.1 as a model for "8-second 720p or 1080p videos featuring stunning realism and natively generated audio" [3]. Clips could be 4, 6, or 8 seconds long at 24 frames per second, one video per request, with 720p as the default; 1080p was limited to 8-second clips [3]. Both 16:9 (the default) and 9:16 were supported for text-to-video and image-to-video [3][19]. Generation from reference images required 8-second clips in 16:9 [3]. Audio was always on [3].
In the Gemini API, Extend worked only on videos previously generated with Veo, at 720p. Each extension added 7 seconds, up to 20 times, for a combined video of up to 148 seconds [3]. The documentation listed request latency of 11 seconds to 6 minutes, said generated videos were stored on Google's servers for 2 days, and said users were not charged if a video was blocked from generating [3]. On Vertex AI, Google said scene extension was "coming soon" at launch; it became available there in preview on December 8, 2025 [2][13].
January 2026 update
On January 13, 2026, Google released an update focused on Ingredients to Video. Google said the updated model "intelligently synthesizes your inputs to preserve character identity and background details, ensuring your characters and settings remain consistent across videos" [7]. The update added three things [7][8]:
- Native vertical Ingredients to Video. Google said Ingredients to Video could now generate "in a native 9:16 aspect ratio" for the first time, producing full-frame vertical video rather than cropping from landscape [7][8].
- 1080p and 4K output. Google's consumer announcement described "1080p and 4K with state-of-the-art upscaling," and its developer post said the model used "state-of-the-art enhancement techniques" to output clearer 1080p and 4K video [7][8]. The Gemini API changelog recorded the change as "Added 4k output resolutions for Veo and more support for portrait videos in all resolutions" [9].
- Wider distribution. Ingredients to Video came to YouTube Shorts and the YouTube Create app for the first time, and the update also rolled out to the Gemini app, Flow, the Gemini API, Vertex AI, and Google Vids. The 1080p and 4K options were limited to Flow, the Gemini API, and Vertex AI [8].
Google said these capabilities, "along with our SynthID digital watermark, are available today in the Gemini API and Vertex AI for enterprises" [7]. After the update, the Gemini API documentation listed 720p, 1080p, and 4K output, with 1080p and 4K restricted to 8-second clips and extension still limited to 720p [3]. Google suggested creating ingredient images with Nano Banana Pro [8].
Availability
Veo 3.1 launched across four surfaces at once. For consumers and creators, it appeared in Flow and in the Gemini app; for developers and enterprises, in the Gemini API (as a paid preview with the model identifier veo-3.1-generate-preview) and in Vertex AI [1][2][3]. 9to5Google reported that Veo 3.1 and Veo 3.1 Fast powered text-to-video and image-to-video in horizontal and vertical formats across the Gemini app, the Gemini API, and Vertex AI [19]. Google included Veo 3.1 in its October 24, 2025 "Gemini Drop" roundup of Gemini app updates, saying users could "create videos with true-to-life textures, easier camera control and dialogue with sound effects" [6]. On Vertex AI, Veo 3.1 and Veo 3.1 Fast entered preview on October 15, 2025 and became generally available on November 17, 2025 [13].
| Surface | Audience | Launch | Status in October 2026 |
|---|---|---|---|
| Flow | Filmmakers, creators | October 15, 2025 [2] | Veo 3.1 Lite, Fast, and Quality still listed alongside Gemini Omni Flash 1.1 [18] |
| Gemini app | Consumers | October 15, 2025 [2][6] | Google says Gemini Omni "will replace Veo in the Gemini app" [17] |
| Gemini API and Google AI Studio | Developers | October 15, 2025, paid preview; model IDs veo-3.1-generate-preview and veo-3.1-fast-generate-preview [3] | Shut down October 22, 2026; recommended replacement gemini-omni-1.1-flash [12] |
| Vertex AI (Gemini Enterprise Agent Platform since April 2026) | Enterprises | Preview October 15, 2025; generally available November 17, 2025 [13] | veo-3.1-generate-001 and veo-3.1-fast-generate-001, retirement date "November 17, 2026 or later" [14] |
SynthID embeds an imperceptible signal into generated video that Google's detection tooling can later read, the same provenance mechanism used across earlier Veo models [3][8].
Veo 3.1 Fast
Google released Veo 3.1 Fast alongside Veo 3.1 on October 15, 2025, in paid preview in the Gemini API under the identifier veo-3.1-fast-generate-preview [3][9]. It supported native audio, first and last frames, and extension like the standard model, but the Gemini API documentation listed reference images (Ingredients to Video) as available for Veo 3.1 only, not Veo 3.1 Fast [3]. Google's documentation described the Fast versions this way: "Veo Fast versions allow developers to create videos with sound while maintaining high quality and optimizing for speed and business use cases" [3]. It suggested them for services that programmatically generate ads, rapid A/B testing of creative concepts, and apps that need to produce social media content quickly [3].
At launch Google said "Veo 3.1 is the same price as Veo 3" [3]. In the Gemini API that meant $0.15 per second of video with audio for Veo 3.1 Fast, against $0.40 for the standard model [10]. Google cut the price of Veo 3.1 Fast on April 7, 2026, to $0.10 per second at 720p, $0.12 at 1080p, and $0.30 at 4K [10][11]. Veo 3.1 Fast became generally available on Vertex AI on November 17, 2025, as veo-3.1-fast-generate-001 [13][14].
Veo 3.1 Lite
On March 31, 2026, Google released Veo 3.1 Lite, which it called its "most cost-effective video model," saying it came "at less than 50% of the cost of Veo 3.1 Fast, but with the same speed" [11]. It supports text-to-video and image-to-video in 16:9 and 9:16, at 720p or 1080p, with clips of 4, 6, or 8 seconds [11]. It was offered on the paid tier of the Gemini API and in Google AI Studio as veo-3.1-lite-generate-preview, priced at $0.05 per second at 720p and $0.08 at 1080p, without 4K output [9][10][11]. Veo 3.1 Lite entered public preview on Vertex AI on April 2, 2026 [13]. Its Gemini API preview endpoint was shut down with the other Veo 3.1 preview models on October 22, 2026 [12]. In Flow, Google's help page says 8-second Veo 3.1 videos can be extended, but only with Veo 3.1 Lite [18].
Pricing
Gemini API prices for Veo 3.1 were per second of generated video with audio. There was no free tier, and Google said users were charged only when a video was generated successfully [10].
| Date | Veo 3.1 | Veo 3.1 Fast | Veo 3.1 Lite |
|---|---|---|---|
| October 15, 2025 (launch) | $0.40 [3][10] | $0.15 [3][10] | n/a |
| After the January 2026 update (4K added) | $0.40 (720p and 1080p), $0.60 (4K) [10] | $0.15 (720p and 1080p), $0.35 (4K) [10] | n/a |
| March 31, 2026 | unchanged [10] | unchanged [10] | $0.05 (720p), $0.08 (1080p); no 4K [10][11] |
| From April 7, 2026 | $0.40 (720p and 1080p), $0.60 (4K) [10] | $0.10 (720p), $0.12 (1080p), $0.30 (4K) [10][11] | $0.05 (720p), $0.08 (1080p) [10] |
| October 22, 2026 | Gemini API preview models shut down [12] | Shut down [12] | Shut down [12] |
At launch prices, an 8-second clip with audio cost $3.20 with Veo 3.1 and $1.20 with Veo 3.1 Fast. In October 2026 Google Cloud's Agent Platform price list still included Veo 3.1, Veo 3.1 Fast, and Veo 3.1 Lite, with the same rates for video with audio as the Gemini API and lower rates for video without audio [15].
Leaderboard standings
Artificial Analysis runs a public text-to-video arena that ranks models by Elo scores computed from blind pairwise human preference votes [21]. On its board for video without audio, an archived snapshot from November 27, 2025, counting all models including retired ones, placed Veo 3.1 Preview fifth with 1,223 points and Veo 3.1 Fast Preview sixth with 1,219, behind Google's own Veo 3 Preview (second, 1,237) and Veo 3 (third, 1,226); first place was held by a model listed as "Whisper Thunder (aka) David" at 1,247 [21]. On January 13, 2026, Runway Gen-4.5 led that board with 1,238, and Veo 3.1 Preview was still fifth with 1,220 among all models (fourth among current models) [21].
By October 2026 the arena had moved to a rescaled board, AA-Video-T2V v2.1, which ranks video generated with audio and fixes Kling 3.0 1080p (Pro) at 1,000 points as its anchor. On it, Veo 3.1 was 22nd with 962 points, Veo 3.1 Fast 23rd with 958, and Veo 3.1 Lite 24th with 953. Google's Gemini Omni Flash 1.1 was 10th with 1,128, and Alibaba's Wan 3.0 led with 1,175 [21].
Notable uses
At launch Google cited two early developer users. Promise Studios, a generative AI film studio, used Veo 3.1 in its MUSE platform for generative storyboarding and previsualization, and Latitude was experimenting with Veo 3.1 in its generative narrative engine to turn user-created stories into video [3]. The January 2026 update put Veo 3.1 Ingredients to Video into YouTube Shorts, the YouTube Create app, and Google Vids [8].
Successor and shutdown
On May 19, 2026, Google announced Gemini Omni, describing it as a model "that can create anything from any input," starting with video, and rolled out its first model, Gemini Omni Flash, to the Gemini app, Google Flow, and YouTube Shorts [16]. Google's Gemini app video page says "Gemini Omni will replace Veo in the Gemini app" and describes Omni as the model "that will now replace the previous Google Gemini Veo 3.1 model" [17].
In the Gemini API, the Veo 3.1 preview models (veo-3.1-generate-preview, veo-3.1-fast-generate-preview, and veo-3.1-lite-generate-preview) were shut down on October 22, 2026, with gemini-omni-1.1-flash listed as the recommended replacement [12]. Google's video generation guide now says: "The Veo 3.1 Preview in the Gemini API has been shut down." It points developers to Gemini Omni Flash or to the generally available Veo models on the Gemini Enterprise Agent Platform [12].
On Google Cloud, Vertex AI became part of the Gemini Enterprise Agent Platform on April 22, 2026 [13]. There the generally available Veo 3.1 and Veo 3.1 Fast models carried a retirement date of "November 17, 2026 or later," and Google listed Veo 3.1 as the recommended upgrade for the Veo 3 and Veo 2 models it retired on June 30, 2026 [14]. Flow continued to offer Veo 3.1 Lite, Fast, and Quality next to Gemini Omni Flash 1.1, though its help page listed Ingredients to Video as unsupported for Veo 3.1 Quality [18].
Reception
Coverage treated Veo 3.1 as an iterative but meaningful update rather than a generational leap. TechCrunch reported Google's claim that "Veo 3.1 builds on May's Veo 3 release and generates more realistic clips and adheres to prompts better," and highlighted the editing additions and the extension of audio across Flow's features [1]. Hassabis promoted it on X as "our new big upgrade with enhanced realism, richer audio, scene extension, better narrative control, more precise editing capabilities & much more" [24].
Coverage framed the release against OpenAI's Sora 2, the successor to its original Sora model. Writing on launch day, Tom's Guide's Amanda Caswell said "if Sora 2 is the TikTok of AI, Veo 3.1 is the Netflix," and argued: "With OpenAI's Sora 2, you get what it gives you. With Veo 3.1, you can shape, tweak and refine what you want to see." [22] The piece described the features Google had announced and noted that the Remove Object feature was "coming soon" [22].
On blind preference votes, Veo 3.1 did not clearly beat its predecessor. In Artificial Analysis's silent text-to-video arena in late 2025 and early 2026, Veo 3.1 Preview ranked below both Veo 3 Preview and Veo 3 [21].
Continue reading
- Veo 3Veo 3 is a video generation model developed by Google DeepMind and announced at Google I/O on May 20, 2025.Mentioned 81 times above
- GoogleGoogle LLC is a United States technology company and the largest operating business within Alphabet Inc. Its activities include internet search, online advertising, video, mobile and web platforms, consumer…Mentioned 56 times above
- Gemini (language model)Gemini is a family of natively multimodal large language models developed by Google DeepMind, first announced on December 6, 2023, that can reason across text, images, audio, video, and code within a single…Mentioned 52 times above
- Google Vertex AIGoogle Vertex AI is the unified machine learning and generative artificial intelligence platform offered by Google Cloud, announced at Google I/O on May 18Mentioned 13 times above
- Google DeepMindGoogle DeepMind is an artificial intelligence research and model-development organization at Google.Mentioned 4 times above
- SynthIDSynthID is a family of digital watermarking technologies developed by Google DeepMind for marking and identifying content generated by generative AI systems.Mentioned 3 times above
More to read
References
- ^1 ^2 ^3 ^4 ^5 ^6 ^7 ^8TechCrunch, "Google releases Veo 3.1, adds it to Flow video editor." October 15, 2025. techcrunch.com/...3-1-adds-it-to-flow-video-editor
- ^1 ^2 ^3 ^4 ^5 ^6 ^7 ^8 ^9 ^10 ^11 ^12 ^13 ^14 ^15 ^16 ^17 ^18Google, "Introducing Veo 3.1 and advanced capabilities in Flow." The Keyword, October 15, 2025. blog.google/...veo-updates-flow
- ^1 ^2 ^3 ^4 ^5 ^6 ^7 ^8 ^9 ^10 ^11 ^12 ^13 ^14 ^15 ^16 ^17 ^18 ^19 ^20 ^21 ^22 ^23 ^24 ^25 ^26Google Developers, "Introducing Veo 3.1 and new creative capabilities in the Gemini API." Google Developers Blog, October 15, 2025. developers.googleblog.com/...ies-in-the-gemini-api ; Google AI for Developers, "Generate videos with Veo 3.1 in Gemini API," archived October 16, 2025: web.archive.org/...video ; archived January 31, 2026: web.archive.org/...video
- ^TechCrunch, "Google's Veo 3 can generate videos and soundtracks to go along with them." May 20, 2025. techcrunch.com/...oundtracks-to-go-along-with-them
- ^Wikipedia, "Veo (text-to-video model)." en.wikipedia.org/...Veo_(text-to-video_model)
- ^1 ^2Google, "Gemini Drops: New updates to the Gemini app, October 2025." The Keyword, October 24, 2025. blog.google/...gemini-drop-october-2025
- ^1 ^2 ^3 ^4 ^5 ^6Google, "Enhanced Veo 3.1 capabilities are now available in the Gemini API." The Keyword, January 13, 2026. blog.google/...veo-3-1-gemini-api
- ^1 ^2 ^3 ^4 ^5 ^6 ^7 ^8Google, "Veo 3.1 Ingredients to Video: More consistency, creativity and control." The Keyword, January 13, 2026. blog.google/...veo-3-1-ingredients-to-video
- ^1 ^2 ^3Google AI for Developers, "Release notes" (Gemini API changelog, accessed October 30, 2026). ai.google.dev/...changelog
- ^1 ^2 ^3 ^4 ^5 ^6 ^7 ^8 ^9 ^10 ^11 ^12 ^13 ^14Google AI for Developers, "Gemini Developer API pricing." Archived snapshots: November 3, 2025 web.archive.org/...pricing ; February 1, 2026 web.archive.org/...pricing ; April 1, 2026 web.archive.org/...pricing ; May 1, 2026 web.archive.org/...pricing ; October 1, 2026 web.archive.org/...pricing
- ^1 ^2 ^3 ^4 ^5 ^6 ^7Google, "Build with Veo 3.1 Lite, our most cost-effective video generation model." The Keyword, March 31, 2026. blog.google/...veo-3-1-lite
- ^1 ^2 ^3 ^4 ^5 ^6 ^7 ^8Google AI for Developers, "Gemini deprecations" (accessed October 30, 2026). ai.google.dev/...deprecations ; "Video generation in the Gemini API" (accessed October 30, 2026). ai.google.dev/...video
- ^1 ^2 ^3 ^4 ^5 ^6 ^7Google Cloud, "Vertex AI release notes" and "Gemini Enterprise Agent Platform release notes" (accessed October 30, 2026). docs.cloud.google.com/...release-notes ; docs.cloud.google.com/...release-notes
- ^1 ^2 ^3 ^4Google Cloud, "Model versions and lifecycle." Gemini Enterprise Agent Platform documentation (accessed October 30, 2026). docs.cloud.google.com/...model-versions
- ^Google Cloud, "Agent Platform pricing" (formerly "Vertex AI pricing," accessed October 30, 2026). cloud.google.com/...pricing
- ^1 ^2Koray Kavukcuoglu, "Introducing Gemini Omni." The Keyword, May 19, 2026. blog.google/...gemini-omni
- ^1 ^2 ^3Google, Gemini app video generation overview (accessed October 30, 2026). gemini.google/...video-generation
- ^1 ^2 ^3 ^4Google Flow Help, "Learn about Google Flow models & supported features" (accessed October 30, 2026). support.google.com/...16352836
- ^1 ^2Abner Li, "Google announces Veo 3.1 and updates Flow with more controls, tools." 9to5Google, October 15, 2025. 9to5google.com/...veo-3-1
- ^Google, "4 ways to refine your content in Flow" (published as "Flow gets new ways to refine and edit videos with AI"). The Keyword, November 25, 2025. blog.google/...flow-refine-videos
- ^1 ^2 ^3 ^4 ^5Artificial Analysis, "Text to Video Leaderboard" (AA-Video-T2V v2.1, accessed October 30, 2026). artificialanalysis.ai/...text-to-video ; archived snapshots: November 27, 2025 web.archive.org/...text-to-video ; January 13, 2026 web.archive.org/...text-to-video
- ^1 ^2Amanda Caswell, "Veo 3.1 is available now (for free) - and it might be the best AI video generator yet." Tom's Guide, October 15, 2025. tomsguide.com/...-its-faster-and-offers-more-tools
- ^Matthias Bastian, "Google's Veo 3.1 update delivers sharper visuals, more realistic textures, and improved audio." The Decoder, October 15, 2025. the-decoder.com/...tic-textures-and-improved-audio
- ^GIGAZINE, "Google releases video generation AI 'Veo 3.1,' supporting video generation of over 1 minute and improved audio quality." October 16, 2025 (quotes Demis Hassabis's post on X). gigazine.net/...20251016-google-veo-3-1
Improve this article
Add missing citations, update stale details, or write a clearer explanation. Anyone can edit: changes go live when you publish, and every edit is checked by an editor.
7 revisions · v7 · 3,311 words · full history
Fact-checks are independent of edits: a reviewer re-verifies the article against its sources and stamps the date. How we verify
Research and drafting on this wiki are AI-assisted, under named human editorial standards. How AI is used here
Reviewer note: Fact-checked on 2026-10-30 against Google's announcements, Gemini API documentation, changelog, pricing and deprecation pages (with archived snapshots), Vertex AI release notes, and archived Artificial Analysis leaderboards, with an independent second review.
Cite this page: AI Wiki. "Veo 3.1." aiwiki.ai, updated 30 Oct 2026, fact-checked 30 Oct 2026. CC BY 4.0. https://aiwiki.ai/wiki/veo_3_1