Google unveils Gemini Omni video generation, replaces Veo 2

Google’s Gemini app now includes “Gemini Omni,” a next‑generation AI video generator that supersedes the Veo 2 model and delivers eight‑second, 720p clips on demand [3].

Gemini Omni replaces Veo 2 in Gemini

Google’s official blog confirms that the Veo 2 model, which produced high‑resolution, cinematic‑style videos, is being retired in favor of Gemini Omni [3]. The new model retains Veo 2’s eight‑second, 720p output but adds “AI‑driven editing” – users can change backgrounds, clothing, or lighting via chat, and an AI avatar can be inserted without re‑uploading images [2]. The rollout is limited to users aged 18+ on the Google AI Plus, Pro, or Ultra plans, with a monthly cap that triggers a warning as the limit is approached [3].

Early leak shows mixed performance

A Reddit user who accessed Gemini Omni through a pop‑up in the Gemini app reported that the model can correctly render logical sequences, such as a professor solving a trigonometric proof on a chalkboard [1]. However, the video exhibited typical AI artifacts: chalk vanished toward the end and hand motions did not always match the writing. A more complex “dinner scene” test revealed objects appearing out of thin air, a known glitch in generative video models. When the user tried the classic “Will Smith eating spaghetti” benchmark, the model refused to generate output, likely due to guardrails [1].

In a side‑by‑side comparison, ByteDance’s Seedance 2 produced a more consistent spaghetti scene, though it suffered from jitter [1]. The Gemini Omni user noted that two video requests already consumed 86 % of their daily quota on the Google AI Pro plan, indicating a tight usage ceiling for early adopters [1].

Competitive context

Veo 2’s specifications—high‑resolution, fluid motion, and physics‑aware rendering—set a high bar for AI video generation [3]. Gemini Omni inherits these capabilities while adding conversational editing, a feature not yet offered by competitors like ByteDance’s Seedance 2. The usage caps, however, may limit its appeal for heavy creators until Google expands the quota.

What to watch

• Quota expansion – Google’s notification system will reveal whether the daily limit is raised for Pro users.
• Official launch timing – Google I/O 2026 may include a formal announcement and broader rollout details.
• Competitor response – Expect ByteDance and other AI video firms to accelerate feature releases or pricing adjustments to counter Gemini Omni’s editing tools.

Google’s shift from Veo 2 to Gemini Omni signals a push toward more interactive, chat‑driven media creation, but the tight usage limits suggest the service is still in a controlled rollout phase. How quickly Google lifts those caps will determine whether Gemini Omni can become a mainstream tool for creators and marketers alike.