Home / Models / Gemini Omni Flash 1.1

Gemini Omni Flash 1.1

Google · Video model

First model of Google's Gemini Omni family, announced at I/O 2026 and built to take any input and, starting with video, produce any output. Version 1.1 Flash reached GA in the Gemini API on 2026-08-27 with video extension, frame interpolation…

Current

ReleasedAug 27, 2026

Context window—

Input → outputtext, image, video, audio → video, audio

API pricegemini-omni-1.1-flash (paid tier): $1.50 / 1M input tokens (text, image, video, audio); output $9.00 / 1M text tokens, $17.50 / 1M video tokens (about $0.10 per second of 720p; 5,792 tokens per second of 720p).

Consumer accessGoogle Flow for Google AI Plus, Pro and Ultra subscribers; scene extension also in the Gemini app for those tiers.

Best for

Conversational video generation and editing where you iterate on a clip in several turns.

Overview

Strengths and limits

Highest-ranked Google video model on Artificial Analysis (Oct 2026)
Accepts text, image, audio and video references; multi-turn editing
360p drafts to 4K output; scene extension and first/last frame

Short base clips (reported up to 10 s; extendable to about 40 s)
Paid tiers only on the API

Published benchmark scores

Artificial Analysis Text to Video Arena Elo (with audio)1132 (rank 8) Source: artificialanalysis.ai

Scores are as published by the source; settings (reasoning effort, tools, attempts) differ between labs, so compare with care. How to read benchmarks →

Sources

Facts checked Oct 14, 2026