Flow Automation - Bulk AI Video & Image Generation

Flow Automation adds batch processing to Google Labs Flow. Instead of entering one prompt at a time and waiting at your screen for each result, you load a list of prompts, choose your settings, and let the extension work through the queue while you do something else. It is built for creators, marketers, and agencies who need more output from Google Labs Flow than manual generation allows. The extension runs in the Chrome side panel next to Flow, so your queue and progress stay visible while you work. WHAT THIS EXTENSION DOES Google Labs Flow is designed around single generations. You write a prompt, wait for the result, download it, and start again. That works well for a few clips. It becomes slow when you need fifty, or two hundred. Flow Automation handles that repetition. You prepare your prompts once, set your model and output options, and start the queue. The extension enters each prompt, waits for the generation to finish, saves the result, and moves to the next task. You can leave the tab and come back when the batch is done. KEY FEATURES Prompt queue Upload a .txt or .csv file containing your prompts, or paste them directly into the panel. Each line becomes a separate task. Tasks are processed in order, and you can pause, resume, or stop at any point. Individual tasks can be removed, or the whole queue cleared, before you start. Five generation modes - Text to Video: a list of written prompts becomes a batch of videos. - Create Image: a list of prompts becomes a batch of images. - Image to Video: upload your own image files and generate video from them. - Ingredients to Video: combine up to three reference images in a single generation, useful for keeping a character or visual style consistent. - Frame to Frame: chain multiple scenes together, where each new scene begins from the final frame of the previous one. Model selection for each task Choose between Veo 3.1, Veo 2, Imagen 4, Nano Banana, and Nano Banana Pro. Fast and Quality tiers are available where Flow offers them, so you can use a faster option while testing ideas and a higher-quality option for final output. Because the model is stored per task, a single queue can mix different models. Output options Select landscape (16:9) or portrait (9:16) aspect ratio, and choose one to four outputs per prompt when you want variations to pick from. Automatic downloads and file naming Turn on automatic download and completed results are saved without manual clicks. Files go into date-based folders inside a base folder you can rename. Filenames are assembled from the prompt text, the model used, the task index, and a timestamp, so files stay distinct and searchable instead of arriving as generic numbered downloads. Optional prefix and suffix fields let you apply your own naming conventions, such as a client code or campaign tag. Error handling for long runs Long batches fail in predictable ways, and the extension is built to keep going: - Each task retries up to three times before being marked as failed. - The page refreshes automatically when errors begin to accumulate. - Repeated failures trigger a pause, so a service problem does not consume the rest of your queue. - Failed tasks are marked and skipped, and the remaining tasks continue processing. Background operation A keep-alive script helps the queue keep running when the Flow tab is not the active tab, so you can use other tabs while a batch is in progress. Timing controls Set the delay between prompts, plus waiting time and stop time values, to control how quickly the extension moves through your queue. Prompt prefix and suffix Apply a fixed piece of text to the beginning or end of every prompt in a run. This is useful for repeating a style instruction across an entire batch without editing each line. Gallery Generated results are stored locally in your browser along with the prompt that produced them. You can filter by media type and re-download items individually or in bulk, including items from earlier sessions. Prompt generator (optional) Expand a list of keywords into longer written prompts using one of six creative presets: General, Cinematic Director, Nature Filmmaker, Abstract Artist, Commercial Director, or Music Video Director. Choose how many prompts to produce per keyword, then send them straight to the queue. This feature requires your own Google Gemini API key, which you enter in the settings. CSV export Export your queue as a CSV file, giving you a record of which prompts were run and what the outcome was. Interface language The panel interface offers a language setting, including an automatic option that follows your system language. HOW IT WORKS 1. Open Google Labs Flow in Chrome, then open the Flow Automation side panel. 2. Add your prompts by uploading a text or CSV file or pasting them in. For image-based modes, select your image files instead. 3. Choose your generation mode, model, aspect ratio, and number of outputs per prompt. 4. Enable automatic download and set your folder and filename options if you want results saved without manual clicks. 5. Press Start. The extension processes each task in order and saves completed results as they finish. WHO IT IS FOR Content creators publishing on a schedule If you post short-form video regularly, the limit is usually production time rather than ideas. Preparing a batch in advance means a single setup session can cover a longer publishing period. Social media and creative agencies When you handle several client accounts, filename prefixes let you tag output by client so files stay separated as they download, rather than needing to be sorted afterwards. Marketers testing creative variations Generating several versions of a concept is straightforward when each prompt can produce up to four outputs and the whole set runs in one queue. Online store owners Image to Video mode works from product photography you already have, which is useful when you want motion content across a range of pr…