Flow AI Video

Flow AI Video is an independent browser-based generator, not a Google product, that gives access to third-party models such as Google's Veo 3 and Veo 3.1, Gemini Omni Video, Kling 3.0, Seedance and Grok Imagine 1.5 Preview. It turns text prompts, still images and frame references into short video drafts. It is aimed at creators, marketing teams, filmmakers, designers, educators and small teams testing shots before production.

A session starts with a choice of mode (image, video or music), a model, an aspect ratio, a clip length and a resolution. The user then writes a shot brief or adds a still image, a first or last frame, or a reference clip. An estimated credit cost appears before submission, and each finished take is kept in the account history for review, download or another pass.

Video modes cover text-to-video, image-to-video, frame-guided shots and reference-led variations in 16:9, 9:16 and 1:1. The Veo 3 page describes 8-second clips with generated sound, up to three reference images and 720p, 1080p or 4K output. Image work runs on Nano Banana, Nano Banana Pro, Nano Banana 2 and GPT Image 2, music on Suno V5, and a remix gallery offers starting prompts.

Usage runs on credits whose cost depends on the model and settings. Plans are Weekly at $4.99 a week, and Pro, Ultimate and Creator at $31, $74 and $139 a month billed yearly, with one-time credit packs also sold. Weekly exports are 720p, higher tiers export 1080p with 2 to 10 concurrent tasks, and credits for failed tasks are returned where supported.