Gemini Omni 1.2: 30% Price Cut, Native 15-Second Video
Google has released Gemini Omni 1.2 Flash, extending its AI video generation capabilities with a 30% price reduction. The model is now available across the Gemini API and enterprise platforms, priced at $0.07 per second for 720p video. Key improvements include support for 15-second clips—up from the prior 10-second limit—with granular duration control in one-second increments, and improved visual consistency across frames and camera transitions.
The model strengthens reference adherence by using images, video clips, and audio to guide generation. Characters, objects, and environments maintain consistency throughout sequences with reduced visual drift. Multi-turn conversational editing enables iterative refinement—swapping backgrounds, adjusting lighting, or changing aspect ratios—without regenerating entire videos, reducing production costs when creating multiple variants.
Additional enhancements target production quality: better lip-sync and dialogue clarity for multi-character scenes, smoother camera movements with fewer jump cuts, and stronger visual consistency in longer sequences. Outputs include imperceptible SynthID watermarking. Development continues on fast-action sequences and complex physics simulations.
Read the full article
Continue reading at Hacker News →This is an AI-generated summary. Read the original for the full story.