Breaking

Google's Gemini Omni Flash Tops Artificial Analysis Video-Generation Leaderboards

July 16, 2026 at 13:08 EDT

Media Generation
Foundation Models

Google DeepMind's newly released Gemini Omni Flash has claimed the top spot on Artificial Analysis's Text-to-Video and Image-to-Video leaderboards, narrowly edging out ByteDance's Seedance 2.0 as of July 2026. The model is the first entry in Google's new Gemini Omni family, which is designed to create and edit high-quality video directly from natural-language descriptions.

July 2026 · Google DeepMind

Gemini Omni Flash Takes the Video Crown

Google's first Gemini Omni model claims the #1 spot on both the Text-to-Video and Image-to-Video leaderboards — narrowly edging out ByteDance's Seedance 2.0. Its edge: multi-turn conversational editing that keeps scene, physics and characters consistent.

On both text-to-video & image-to-video boards

720p

Resolution at 24 fps, 16:9 or 9:16

8.4/10

Hands-on review rating

Per minute via API ($0.10/sec)

Clip length ceiling

Outputs run only 3–10 seconds — great for short-form, short for narrative.

minimum

10s

maximum

The differentiator: multi-turn conversational editing

Generate clip from prompt

→

Refine via chat

→

Keep scene, physics & character consistent

Stateful editing preserves history via the Interactions API. Native audio (speech, music, SFX) plus an automatic SynthID watermark on every output.

Strengths

Fast, practical, strong iterative editing
Ads, product spots, vlogs, explainers
Replaces Veo 3.1 in the Gemini app

Limitations

10-second ceiling limits narrative
Uneven prompt adherence; motion failures
Strict policy triggers unexpected refusals

The margin over Seedance 2.0 is slim — and the race is tightening fast.

Google, ByteDance, Kling, Runway and Luma are all closing in. By folding video into the Gemini stack across consumer apps, developer tools and enterprise, Google is betting conversational editing goes mainstream — even as longer-form, cinematic quality stays an open challenge.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.