Published 3 days ago • loading... • Updated 2 days ago
Gemini Omni Will Bring Only More AI Slop and Skepticism
Google says the model can turn photos and selfies into realistic clips and will roll out through Gemini, Flow and YouTube Shorts today.
On Wednesday, Google unveiled Gemini Omni, a generative AI model that creates realistic videos from text, images, and audio inputs. The tool is available today via Gemini App, Google Flow, and YouTube Shorts.
Developing Omni as a "world model," the company has spent two years playing catch-up in generative video. The model simulates real-world physics like gravity and fluid dynamics, bridging photorealism with meaningful storytelling.
Users can "edit the action, add in new characters or objects" through natural conversation, with each instruction building on previous turns. Omni maintains character and environment consistency across edits, enabling creative refinement.
To address safety concerns, Google implemented "clear policies to protect users from harm and governing the use of our AI tools." All generated videos receive an invisible SynthID digital watermark for identification.
Gemini Omni Flash is available now for AI Plus, Pro, and Ultra subscribers starting at $7.99 per month, with free access on YouTube Shorts. Google teased a higher-level Omni Pro model with details coming soon.
Provide a photo, a few words, an audio excerpt — Gemini Omni does the rest. Unveiled to Google I/O 2026, Tuesday, May 19, this multimodal IA generates consistent and visually credible videos from...