Reported with AI assistance from public sources. Every claim names its outlet in the text and is listed with its source below. An editor pass and fact-check ran before publication. — Align Newsroom, part of Creytix.
DeepMind has released Gemini Omni 1.2 Flash, a model described as fast and cost-effective with longer clips, according to the DeepMind blog .
The release follows DeepMind's announcement of its next frontier model in late October , which an AI industry newsletter said topped benchmark charts, for now .
Speed, cost and clip length are the three measures the release leads with . A model that is faster and cheaper while handling longer clips changes the economics of generating video, which is the use the longer clips point to.
What is known is the release itself and the three claims made for it: fast, cost-effective, longer clips . What is not yet known is how the model performs against rivals, what it costs to use, or when it will be widely available — the announcement in today's intake gives no benchmark figures or prices. The release was announced on DeepMind's own blog, and no independent outlet in today's intake had confirmed it .
Sources
- Google DeepMind — Blog Single source
- The Rundown AI — Newsletter
Single source means the only outlet cited for at least one claim in this article.