Google Opens Gemini Omni Video Generation to Free Vids Users After Paid-Only May Launch
Key Takeaways
- •Google made Gemini Omni 1.1 Flash video generation free for all Google and Workspace account holders inside Google Vids starting September 23, after restricting it to paid AI tiers since its May launch.
- •The free release introduces three new controls—consistent scene extension, duration locking tied to scripts or narration, and 1080p upscaling of older AI footage—and all are available to non-paying users.
- •Free access comes with output limits, as additional generation on personal accounts requires a subscription, while Workspace Business and Enterprise customers receive larger admin-managed pools.
- •A hands-on test produced 720p clips instead of the 1080p advertised in Google's blog, though one short clip was completed in under 30 seconds.
- •Every Omni 1.1 clip created in Vids embeds an invisible SynthID watermark that Google's SynthID Detector can read to confirm a video was AI-generated.
Google removed the paywall on Gemini Omni video generation in Google Vids on September 23, allowing anyone with a Google or Google Workspace account to produce 1080p clips at no cost using the Gemini Omni 1.1 Flash model. The change makes the model available to non-subscribers inside Vids for the first time since its debut.
The feature had been restricted to subscribers of Google's AI Plus, Pro, and Ultra tiers since its launch in May. Non-paying users could previously access it only through YouTube Shorts and YouTube Create. With the paywall lifted, Vids now stands alongside those apps as a free route to the model.
As detailed in Google's official blog, the free release introduces three new controls. A scene can be extended without the characters, lighting, or setting shifting mid-way. Creators can lock a clip to a specified duration that aligns with a script or narration. In addition, older AI footage is upscaled to 1080p rather than regenerated. All three controls ship to free users rather than being reserved for paying tiers.
Omni 1.1 Flash builds video from text prompts, images, audio, or existing clips. Additional output on a personal account requires a Google AI subscription, while Workspace Business and Enterprise customers receive larger pools that administrators manage centrally. Free access therefore comes with an output ceiling that a subscription raises.
The tool is available at vids.new, where desktop users can click "Create AI videos" to begin. "We integrated our newest AI models into Google Vids so anyone with a Google Account can create videos for free," wrote Justin Luk, a product manager on the Google Vids team.
A hands-on test produced results that fell short of the full 1080p advertised in Google's blog: the clips generated were 720p, in both horizontal and vertical formats, though one short clip was completed in under 30 seconds. The gap between the advertised spec and the observed output is something users can check for themselves when trying the free tier.
Google also plans to incorporate Gemini 3.8 Flash-Lite text-to-speech, which will read typed scripts aloud in more than 100 languages.
Every Omni 1.1 clip created in Vids carries SynthID, a Google DeepMind watermark embedded in the video frames. The mark is invisible to viewers but readable by software. Google's SynthID Detector can identify the watermark, allowing anyone who uploads a clip to the tool to verify that it was made with AI. Because every clip is marked, output from free users can be checked the same way as output from subscribers.
