FLOW · Auto Veo, Omni & Nano Banana on Google Flow

Flow Automation for Google Flow: auto Veo & Omni AI videos, Nano Banana images, prompt batches and a local API for AI agents.

FLOW is API and AI agents automation for Google Flow. Queue AI video and image generations, run whole prompt files in one click, and drive Google Flow from your own scripts and AI agents through a local API — all inside your own Chrome, with your own Google account. AI VIDEO • Omni 1.1 Flash: text to video, first frame, first + last frame, and Ingredients (up to 7 reference images); 4–10 s • 720p • Veo 3.1 Lite and Veo 3.1 Fast from a first frame • 1–4 variants per prompt, 16:9 or 9:16 AI IMAGES • Nano Banana 2.1, Nano Banana 2 Lite and Nano Banana Pro • 1–8 images per prompt, 5 aspect ratios, 1K / 2K / 4K upscale, PNG, JPEG or WEBP • Up to 10 reference images PROMPT BATCHES • Drop a .txt file: every prompt becomes its own job, with reference images by path or link • Name a result at the end of a prompt — ;[Max] saves Max.png — and reuse it in later prompts as @Max: a consistent character from image to video • Files are saved to any folder you choose, named after the prompt file (Story_0001.png) LOCAL API FOR AUTOMATION • HTTP API at http://flow.local:7777 for curl, Python, JavaScript, n8n, Make and other no-code tools • Async jobs with status and file download, access tokens, everything stays on your computer • LLM.md: the whole API in one document for ChatGPT, Claude and other AI agents — paste it and the agent starts generating ACCOUNTS AND QUEUE • Several Google accounts, switch in one click; the queue moves on to an account with credits left • Default settings for video and images, a job list with previews, retry and reuse