{
"$type": "site.standard.document",
"bskyPostRef": {
"cid": "bafyreih4tna7s46em2gxkvkemscjgrkb2bvxo6uj4awb4zfvy5thrpuhgq",
"uri": "at://did:plc:dz7fbvkxedbwlm4sroohfpee/app.bsky.feed.post/3mkvozv4sczl2"
},
"coverImage": {
"$type": "blob",
"ref": {
"$link": "bafkreif3ehmphfrehtji56sxfk2okmxwp55jjiej5mq4tkwcj4lvzbpeky"
},
"mimeType": "image/jpeg",
"size": 122545
},
"description": "Google may introduce a new Gemini video-generation tool called Omni, as hinted by new UI leaks ahead of I/O 2026.",
"path": "/google-is-testing-new-omni-model-for-video-generation-ahead-of-i-o/",
"publishedAt": "2026-05-02T22:30:41.000Z",
"site": "https://www.testingcatalog.com",
"tags": [
"ahead of Google I/O 2025",
"Source"
],
"textContent": "Google appears to be preparing a new Gemini video-generation tool called **Omni**. A recently surfaced screenshot from Gemini’s video generation tab includes the line, **“Start with an idea or try a template. Powered by Omni.”** The placement matters because Omni appears near where \"Toucan\" is mentioned, a currently active video generation tool powered by Veo, spotted ahead of Google I/O 2025.\n\nToday, Gemini’s video generation flow is presented as powered by Veo 3.1, while image generation is tied to Nano Banana 2 and Nano Banana Pro, with Google describing Nano Banana Pro as built on Gemini 3 and Nano Banana 2 as Gemini 3.1 Flash Image. The open question is whether Omni is a new wrapper for Veo, a new Gemini video model, or an early step toward a Gemini omni-model that can handle images and videos within a single system. Since Omni appears in a visible UI string, not only in hidden references, it may also be intended as a public product name.\n\n> **If true (as it is still highly speculative), Gemini will be the first top-tier Omni model with video output!**\n\nGoogle currently runs a split-model strategy, with Veo for video and Gemini-based Nano Banana models for image generation. Omni could bring those tracks closer together. It would also arrive amid competition in AI video, with ByteDance’s Seedance 2.0 topping video-gen benchmarks.\n\nThe likely launch window to watch is Google I/O 2026. Google says the event runs May 19–20 and will include Gemini and broader AI updates, making it a plausible stage for a larger Gemini media-generation reveal.\n\nSource",
"title": "Google is testing new Omni model for video generation ahead of I/O",
"updatedAt": "2026-05-02T22:30:41.677Z"
}