How to Use AutoBoom

Give your project a name. This is just for you — it helps you find it later in Saved Projects and History. Something like "Beach Sunset Scenes" or "Product Showcase Batch".

Choose between 9:16 (Portrait) or 16:9 (Landscape). This applies to all images and videos in the project. Pick 9:16 for TikTok/Reels, 16:9 for YouTube.

How many images or videos to generate for each prompt. Set to 1 for speed, or up to 4 if you want variations to pick from.

Pick one of the three modes: Text to Video, Frames to Video, or Create Image. Each mode has its own screen — explained in detail below.

Hit the "Continue" button to go to your selected mode. You can always go back to the home screen later.

Click "+ Add Image Prompt" to add a prompt field. Type what you want the image to look like. Be specific — describe the scene, lighting, style, colours.

Add as many prompts as you want. Each prompt creates its own image.

Got a long list of prompts? Click "📋 Bulk Prompts" PRO to paste them all at once. Separate each prompt with a blank line. Hit "Import Prompts" and they'll each become their own prompt field.

Hit the "🎨 Create Images" button. AutoBoom takes over from here. It goes to Google Flow, types each prompt, clicks Create, waits for the result, then moves to the next one.

You'll see a progress bar and status updates: which prompt is being generated, what phase it's in, and how many are done. You can Pause or Stop the run at any time.

When all images are generated, you'll see a completion card. Click "Done — New Project" to start fresh, or go to History to review what you made.

At the top you'll see two tabs: Single and Batch.

Single: One prompt, one video. Good for testing a prompt or making a quick clip.

Batch: Multiple prompts, each on its own line. AutoBoom generates a video for each one, back-to-back.

Describe the video you want. Be descriptive — include the subject, motion, camera angle, mood, and style.

Example: "A cinematic drone shot flying over snow-capped mountains at golden hour, with clouds drifting slowly below."

For batch mode, put each prompt on its own line.

Hit the "🚀 Create Videos" button. AutoBoom navigates to the Text to Video tool on Flow, enters your prompt, picks the right model and aspect ratio, and submits it.

Videos take longer than images — usually a few minutes each. AutoBoom shows you the progress with a bar and prompt dots. Each dot represents one prompt.

Once all prompts are submitted, AutoBoom waits for Google Flow to finish rendering them.

When videos are ready, the "📥 Download Videos" button lights up. Click it to download all your videos as a ZIP file. Or click "Done — New Project" to move on.

These are your "keyframes" — the images that will become the starting and ending points of each video clip. Click "+ Add Image Prompt" and describe each scene.

Example: You want a video that goes from a sunrise to a sunset. You'd add two image prompts — one for the sunrise, one for the sunset.

These describe the motion between your images. Each animation prompt creates a video transition between two consecutive images.

If you have 5 images, you normally need 4 animation prompts (one between each pair). Or if Chain Mode is off, each image gets its own animation prompt.

Example: "Slow zoom out as the sun rises over the horizon, light gradually warming the scene."

Click "Save Project" to save it locally, then click "▶ Run" to start. AutoBoom goes through two phases:

Phase 1 — Images: AutoBoom generates all your images on Flow, one by one. It grabs the URL of each generated image for the next phase.

Phase 2 — Videos: AutoBoom switches to the Frames-to-Video tool on Flow. For each pair of images, it uploads the start frame and end frame, enters the animation prompt, and submits the video. Then it waits for all videos to render and downloads them.

The progress screen shows you which phase you're in (Images or Videos), which step it's on, and a big progress bar. You can Pause, Resume, or Stop at any time.

Once all videos are rendered, the "Download Videos" button activates. Click it to get all your videos in a ZIP. Your images are already saved on Flow — you can find them there too.

Turn on the 🔗 Chain Mode toggle at the top of the Create Image screen. Now, when AutoBoom generates Image 1, it takes the URL of that result and feeds it as a reference into Image 2. Image 2's result feeds into Image 3, and so on.

This means every image builds on the last one. Great for creating a series of images that look like they belong together — same character, same style, same world.

You can also set a "First Image Reference URL" if you already have a starting image you want to use.

Chain Mode changes how video transitions work. With Chain Mode on, AutoBoom uses consecutive pairs: Image 1 → Image 2, Image 2 → Image 3, and so on. Each pair becomes one video clip.

With Chain Mode off, each image stands alone — it gets its own animation prompt and its own video.

In Create Image or Frames to Video mode, you'll see an "✨ AI Prompt Parser" section. Click the "Paste All" button to open the paste area.

Paste your full prompt document into the text area. This can be a ChatGPT output, a script, a list — any format. It doesn't need to be perfectly structured.

The AI will figure out which parts are image prompts and which parts are animation prompts.

Hit the "🤖 Parse with AI" button. AutoBoom sends your text to an AI model, which reads through it and extracts the individual prompts. You'll see a status spinner while it works.

The parsed prompts fill into the fields automatically. Scroll through them and edit any that need tweaking. Then you're ready to run.

In any mode (Create Image, Text to Video, or Frames to Video), you'll see a "+ Add to Batch" button. Click it to save the current project to the batch queue instead of running it immediately.

Go back to the home screen and click "Batch Projects". You'll see all your queued projects listed. You can drag-and-drop to reorder them, select/deselect with checkboxes, and remove any you don't want.

Set the cooldown between projects (default is 30 seconds — this gives Flow a breather). Set the error limit — how many consecutive failures before AutoBoom stops the whole batch.

Click "▶ Start All" to run every project in order, or "▶ Start Selected" to run only the checked ones. AutoBoom runs them back-to-back until the queue is done.

Click the 👻 ghost icon in the top-right of the AutoBoom panel. A dark overlay covers the Flow page. AutoBoom still runs underneath — it just hides the browser from view.

The overlay shows a small progress indicator so you still know what's happening. Click the ghost icon again to turn it off.