Automate Google Flow Image & Video Generation — All Prompts at Once (You Need This)

Summary

The video demonstrates AutoFlow, a Chrome extension for automating image and video generation in Google Flow. Instead of entering prompts and selecting references one at a time, users can submit batches of prompts and let the extension process them, generate media, and download and rename the resulting files.

Google Flow overview

The presenter describes Google Flow as offering:

• Image generation: Multiple aspect ratios and three image models, identified in the subtitles as Nano Banana 2.1, 2 Lite, and Pro. The presenter says Flow can generate up to four images at a time.
• Video generation: Vertical or horizontal output, reference images and start frames, and several video models. Video lengths are described as 4–10 seconds.
• Credits: The presenter says a 10-second video uses 15 credits and that the free allowance is 50 credits—about three such videos per day. He also promotes a paid Google AI Pro plan, claiming it includes monthly credits and additional storage.

These product limits and plan details are the presenter’s claims and may change.

Installing and configuring AutoFlow

The tutorial instructs viewers to download the extension as a ZIP file, extract it, and install it in Chrome through Manage Extensions:

1. Turn on Developer mode.
2. Choose Load unpacked and select the extracted extension folder.
3. Keep the folder in place; deleting it will remove the extension.

In the extension’s settings, the presenter recommends adjusting the processing speed and delay, enabling retries, choosing a download quality and save folder, and turning on automatic file renaming. He advises saving the settings before starting a workflow.

Automated workflows

AutoFlow provides six controls: Text to Image, Image to Image, Text to Video, Frame to Video, Ingredients to Video, and Agent Mode.

• Text to Image: Paste multiple prompts separated by a blank line, select the format and model, and start AutoFlow. The presenter demonstrates batching up to 100 prompts. The extension processes them sequentially and downloads the results.
• Character consistency: Upload character images and give them names that match the names used in prompts. The presenter demonstrates scanning characters from Flow so the extension can associate them with relevant scenes.
• Image to Image: Upload reference images, give them clear names, and mention those names in the prompts. The presenter says to enable automatic reference selection and upload the images to Flow’s Add ingredients area. The extension then selects characters and references for each prompt.
• Text to Video: Paste a batch of video prompts and scan for characters. The presenter also shows how to assign character voices in advance to keep them consistent when characters speak.
• Frame to Video: Add start and end images, then provide prompts describing how to animate the scene.
• Ingredients to Video: Add character and reference assets, then submit prompts for scenes. The presenter describes this as similar to Image to Image, but for generating videos.
• Agent Mode: The presenter says it is similar to Google Flow’s built-in agent automation but skips a full demonstration.

Prompts should clearly identify characters and reference images. The presenter warns that ambiguous or mismatched names can prevent automatic selection from working.

Additional points

The video focuses on automating generation rather than editing. It briefly advertises a future tutorial about automating voice-over syncing, motion, keyframes, transitions, and subtitles; those editing steps are not demonstrated.

Main speaker/source: A single presenter from AIWaliSarkar, demonstrating Google Flow and the AutoFlow extension. The presenter says he created the extension and promotes a tool provider for paid-plan access.