Veo 3.1 is the next update to Google’s AI video model. It is showing up in partner docs and previews with claims of longer clips, 1080p output, better physics, and new creative controls. Google has not posted an official Veo 3.1 release bulletin yet. Treat anything beyond Veo 3 as in preview or rumor until Google publishes final notes.
What is Veo 3.1
Veo is Google DeepMind’s text to video model. The current public docs focus on Veo 3, which already supports native audio generation for sound effects, ambient noise, and even dialogue. Veo 3 also ships inside Google’s developer stack and select apps. Veo 3.1 is described by partners as the next step with higher fidelity and more control.
How to try
Higgsfield early access
Go to higgsfield.ai/veo3.1. Click Join Waitlist to request early access to Veo 3.1 inside the Higgsfield app. The page shows the Veo 3.1 intro and a visible Join Waitlist button.
What is new compared to Veo 3
What Google has documented for Veo 3 today
Veo 3 preview on Vertex AI supports 720 and 1080 resolution, with video durations of 4, 6, or 8 seconds per generation through the API. Google added 1080 upscaling support in July. You can also extend a Veo clip inside Vertex AI to go beyond the first segment.
What outside pages say about Veo 3.1
- Longer clips than 8 seconds, with some pages claiming 10 seconds today and talk of 30 seconds or even 1 minute in 1080p. Take these as early partner claims until Google publishes specs.
- Better physics and scene control. Some partners mention start and end frame control and multi shot storytelling.
- Automatic sync for audio, voiceover, and effects inside one generation. This is listed on partner preset pages.
Bottom line
Right now, the only durations and limits you can rely on are the ones in Google’s own docs for Veo 3. Everything else sounds promising but stays unconfirmed until Google updates the official pages.
Status and timing
TestingCatalog spotted Veo 3.1 references in Vertex AI, Google Vids, and the Higgsfield platform, which suggests a staged rollout. A separate roundup notes there is still no official 3.1 bulletin from Google. Some blogs tied the rumors to early October events, but no formal date was posted.
How you can try Veo right now
You can work with Veo 3 today while Veo 3.1 firms up.
- Vertex AI Studio or API
Generate 4 to 8 second clips at 720 or 1080, then use the extend option to build longer sequences. This is the most direct path for developers. - Platforms that integrate Veo
Higgsfield shows Veo support and has public pages that mention 3.1 presets and multi shot tools. This is useful for creators who want a UI without code. Treat any 3.1 tagged features as preview until Google confirms. - Google apps and labs
Google experiments like Flow and Vids have featured Veo 3. Official coverage also shows Veo 3 landing in consumer surfaces over time. These are helpful for non technical users.
Features to watch for in Veo 3.1
- Longer runtime in HD. Pages mention 10 seconds now and talk of 30 seconds to 1 minute later. Wait for Google’s spec before you plan production.
- Stronger consistency and physics. Partners call out better character consistency and physical realism. Veo 3 already focuses on realism and prompt adherence, and 3.1 is expected to push that further.
- Creative controls. Start and end frame control and multi shot scenes for story structure. Good for ads, shorts, and B roll.
- Audio in the same pass. Partner pages claim automatic sync across dialogue, music, and effects. Veo 3 already generates audio natively.
FAQ
Is Veo 3.1 officially released
No official release bulletin is posted yet. Partners and trackers report sightings and previews.
Does Veo 3 already support 1080
Yes. Veo 3 supports 720 and 1080 in preview, and Google added 1080 upscaling support in July.
Can Veo generate sound
Yes. Veo 3 can generate sound effects, ambient audio, and even dialogue. Partners say 3.1 improves sync inside one render.
Will Veo 3.1 do 30 seconds or 1 minute
Some pages claim that. Treat it as unconfirmed until Google posts specs.