User Guide – CreatorFlow

🚀 Getting Started ▼

Open labs.google/fx/tools/flow and navigate to a Flow project page.

Click the extension icon in Chrome's toolbar and pin it for easy access.

Open the Control tab in the side panel and select your desired mode.

Configure settings in the Settings tab before your first run.

⚠ Important

The extension only works when you are on a Flow project page. If the panel shows "not active", navigate to a project and refresh.

🎛️ Modes Overview ▼

Mode 1

Text-to-Video

Generate videos from text prompts. Separate prompts with blank lines.

Mode 2

Image-to-Video

Animate uploaded images with text prompts. 1–2 images per prompt.

Mode 3

Components-to-Video

Use character/component images with auto filename matching. Up to 3 images per prompt.

Mode 4

Text-to-Image

Generate images from descriptions. Supports 4K/2K/1K download.

Mode 5

Image-to-Image

Transform source images with prompts. Auto filename matching. Up to 10 images per prompt.

📋 Mode Details ▼

Mode 1 — Text-to-Video

Enter prompts in the text area, separated by blank lines, or upload a .txt file.
Configure duration: 8s or Concat.
Click Run to start batch generation.

Mode 2 — Image-to-Video

Upload images (supports multiple at once via file picker or drag-drop).
Enter one prompt per image (or per image pair).
Sort images by name or date, or reorder via drag-and-drop.
Configure 1–2 images per prompt in Settings.

Mode 3 — Components-to-Video

Upload component or character images.
Enter prompts describing the scene.
Enable "Auto-add character images" — the tool matches images to prompts by filename. For example, dog.png and cat.png are automatically attached when the prompt mentions "dog" and "cat".
Up to 3 images per prompt.

💡 Tip

Name your images clearly and descriptively (e.g. hero.png, dragon.jpg) so auto-matching works reliably.

Mode 4 — Text-to-Image

Enter image descriptions, one per blank-line-separated block.
Configure aspect ratio and image model in Settings.
Supports 4K / 2K / 1K auto-download.

Mode 5 — Image-to-Image

Upload source images and enter transformation prompts.
Enable "Auto-add character images" for filename-based matching.
Up to 10 images per prompt.
Supports 4K / 2K / 1K auto-download.

⚙️ Settings Reference ▼

Setting	Options	Notes
Default Mode	Mode 1–5	Mode loaded on panel open
Default Aspect Ratio	16:9 · 9:16 · 1:1	For image generation modes
Outputs per Prompt	1 – 4	Videos or images generated per prompt
Concurrent Prompts	1 – 6	How many prompts run in parallel
Prompt Delay	0 – 300 s	Wait between prompt submissions
Video Model	Veo 3.1 Fast · Veo 2	Quality vs. speed tradeoff
Image Model	Various	Model used for image generation
Auto Download Video	720p · 1080p · No Download	Automatic save after generation
Auto Download Image	1K · 2K · 4K · No Download	Automatic save after generation
Max Retries	1 – 20	Retries on generation failure
Default Video Frame	8s · Concat	Duration per video clip
Language	English · Vietnamese · Chinese	UI language

💡 Tips & Best Practices ▼

Use English prompts for best generation quality.
Separate prompts with a blank line — each block is one prompt.
Set ~3 concurrent prompts when Flow is stable; drop to 2 during peak hours.
Set a ~30 s delay between prompts to avoid rate limiting.
Name images clearly for auto-matching: dog.png, hero.jpg, etc. Supported formats: PNG, JPG, GIF (under 10 MB).
Lower concurrency = more stable runs with fewer errors.
Use Concat duration to chain multiple clips into one long video.

🔧 Troubleshooting ▼

❓ Extension not working / panel shows "not active"

You must be on a Flow project page at labs.google/fx/tools/flow. Navigate to a project and refresh the page.

❓ Cannot create video

The tool auto-retries every 30 s. This is normal — just wait and don't close the panel.

❓ Videos not downloading automatically

Go to Chrome Settings → Downloads and turn OFF "Ask where to save each file before downloading".

❓ Policy Error on a prompt

The prompt or image violates Google's content policy. The tool automatically skips that prompt and continues with the rest.

❓ Screen zoomed out unexpectedly

This is an automatic feature of the extension. Do not manually adjust zoom while a run is in progress.

🚀 Getting Started ▼

Open grok.com/imagine and make sure you are logged in to Grok.

Click the extension icon and pin it to the toolbar.

In the Control tab, configure: Concurrent Prompts, Prompt Delay, Save Folder, and Auto Change File Name.

Select your mode and click Run.

⚠ Important

The extension only activates on grok.com/imagine. If the panel shows "not active", navigate to that URL and refresh.

🎛️ Modes Overview ▼

Mode 1

Text-to-Video

Generate videos from text prompts separated by blank lines.

Mode 2

Frame-to-Video

Animate images as start frame, or start + end frame pair.

Mode 3

Ingredients to Video

Character images auto-matched by filename. Up to 10 images per prompt.

Mode 4

Text-to-Image

Generate images from descriptions with configurable aspect ratio.

Mode 5

Image-to-Image

Transform source images with prompts. Auto filename matching.

📋 Mode Details ▼

Mode 1 — Text-to-Video

Enter prompts separated by blank lines, or upload a .txt file.
Click Run to start batch generation.

Mode 2 — Frame-to-Video

Upload or drag-drop source images.
Enter prompts for each image.
In Settings → Image Processing Option: choose Start frame only or Start + End frame.
Make sure you have enough images for the number of prompts.

Mode 3 — Ingredients to Video

Upload component or character images.
Enter prompts describing the scene.
Enable "Auto-add character images" for automatic filename-based matching.
Settings → Max Input Images per Prompt: 1–10.

💡 Tip

Name character images descriptively: hero.png, villain.jpg, dragon.png. The tool matches image filenames to words in your prompts.

Mode 4 — Text-to-Image

Enter image descriptions, one per blank-line-separated block.
Configure Aspect Ratio and Image Model in Settings.

Mode 5 — Image-to-Image

Upload source images and enter transformation prompts.
Enable "Auto-add character images" for filename-based matching.
Settings → Max Input Images per Prompt: 1–10.

Video Chaining with Extend

💡 Pro Tip

Use 6s extend or 10s extend as the video option for all prompts except the last one. This chains them into one continuous long video. Set the last prompt to a non-extend option.

⚙️ Settings Reference ▼

Setting	Options	Notes
Default Mode	Mode 1–5	Mode loaded on panel open
Default Aspect Ratio	16:9 · 9:16 · 1:1 · 2:3 · 3:2	For image generation modes
Outputs per Prompt (Video)	1 – 4	Videos generated per prompt
Outputs per Prompt (Image)	1 – 50	Images generated per prompt
Concurrent Prompts	1 – 6	Parallel prompt submissions
Random Delay	Toggle + range	Adds randomness to avoid rate limits
Video Model	Grok 3.1 Fast · Grok 3.1 Quality · Grok 2	Quality vs. speed tradeoff
Image Model	Various	Model used for image generation
Default Video Option	6s · 10s Super Grok · 6s extend · 10s extend	Duration and chaining mode
Default Image Mode	New Image · Edit Image	Create new or edit existing
Image Processing Option	Start frame only · Start + End frame	For Frame-to-Video mode
Max Input Images per Prompt	1 – 10	For Ingredients/Image-to-Image modes
Max Retries	1 – 20	Retries on generation failure
Auto Download Video	No Download · 480p · 480p Upscale · 720p Super Grok	Automatic save after generation
Auto Download Image	No Download · 1K	Automatic save after generation
Language	EN · VI · ZH · KO · ES · JA	UI language

💡 Tips & Best Practices ▼

Increase Random Delay if you hit rate limits frequently.
Start with 1 concurrent prompt and increase slowly to find your stable limit.
Use 6s extend / 10s extend to chain prompts into one long video — set the last prompt to a non-extend option.
Use the Quick Fix button if generation gets stuck.
Name character images descriptively for auto-matching: hero.png, villain.jpg.
For Frame-to-Video, ensure you have at least as many images as prompts.

🔧 Troubleshooting ▼

❓ Extension not active

Must be on grok.com/imagine. Navigate there and refresh the page.

❓ Generation errors

The tool auto-retries. Use the Fix Error button if it stays stuck.

❓ Downloads not working

Go to Chrome Settings → Downloads and turn OFF "Ask where to save each file before downloading".

❓ Login required error

You must be logged into Grok.com before using the extension.

❓ Not enough images for Frame-to-Video

Frame-to-Video requires at least one image per prompt. Upload enough images to match your prompt count.

🚀 Getting Started ▼

Open dreamina.capcut.com (or jimeng.jianying.com) and log in.

Click the extension icon and pin it to the toolbar.

Open the Control tab and configure your settings.

Select your mode and click Run.

⚠ Important

The extension only activates on dreamina.capcut.com or jimeng.jianying.com. If the panel shows "not active", navigate to one of those sites and refresh.

🎛️ Modes Overview ▼

Mode 1

Text-to-Video

Generate videos from text prompts. Blank line separator.

Mode 2

Frame-to-Video

Animate images as start frame or start + end frame pair.

Mode 3

Ingredients to Video

Character images auto-matched by filename. Up to 10 images per prompt.

Mode 4

Text-to-Image

Generate images from descriptions. Expanded aspect ratio options.

Mode 5

Image-to-Image

Transform source images. Supports image chaining in Edit Image mode.

📋 Mode Details ▼

Mode 1 — Text-to-Video

Enter prompts separated by blank lines, or upload a .txt file.
Configure duration: 4s–15s or Concat variants.
Click Run to start batch generation.

Mode 2 — Frame-to-Video

Upload or drag-drop source images.
Enter prompts for each image.
Settings → Image Processing Option: Start frame only or Start + End frame.

Mode 3 — Ingredients to Video

Upload component or character images.
Enter prompts describing the scene.
Enable "Auto-add character images" for filename-based matching.
Settings → Max Input Images per Prompt: 1–10.

Mode 4 — Text-to-Image

Enter image descriptions, one per blank-line-separated block.
Configure Aspect Ratio (includes 4:3 and 3:4) and Image Model in Settings.

Mode 5 — Image-to-Image (with Image Chaining)

Upload source images and enter transformation prompts.
Enable "Auto-add character images" for filename-based matching.
Enable Edit Image mode to activate Image Chaining: each prompt's output image automatically becomes the input for the next prompt.

💡 Image Chaining

Image Chaining is a Seedance-exclusive feature. Enable Edit Image mode in Settings to chain prompts — great for iterative style transformations or storytelling sequences.

Video Chaining with Concat

💡 Pro Tip

Use Concat duration to chain multiple video clips into one long video. Each prompt generates a clip that is appended to the previous one.

⚙️ Settings Reference ▼

Setting	Options	Notes
Default Mode	Mode 1–5	Mode loaded on panel open
Default Aspect Ratio	16:9 · 9:16 · 1:1 · 4:3 · 3:4 · 2:3 · 3:2	Expanded vs. Grok
Outputs per Prompt (Video)	1 – 4	Videos generated per prompt
Outputs per Prompt (Image)	1 – 50	Images generated per prompt
Concurrent Prompts	1 – 6	Parallel prompt submissions
Random Delay	Toggle + range	Adds randomness to avoid rate limits
Video Model	Various Seedance models	Quality vs. speed tradeoff
Image Model	Various	Model used for image generation
Default Video Duration	4s · 5s · 8s · 10s · 15s · Concat variants	Duration per video clip
Default Image Mode	New Image · Edit Image	Edit Image enables image chaining
Image Processing Option	Start frame only · Start + End frame	For Frame-to-Video mode
Max Input Images per Prompt	1 – 10	For Ingredients/Image-to-Image modes
Max Retries	1 – 20	Retries on generation failure
Auto Download Video	No Download · 720p · 2K · 720p 30FPS · 2K 30FPS · 720p 60FPS · 2K 60FPS	Highest quality: 2K 60FPS
Auto Download Image	No Download · 1K · 2K	Automatic save after generation
Language	EN · VI · ZH · KO · ES · JA	UI language

💡 Tips & Best Practices ▼

Use Concat duration to chain prompts into one long video.
Use Edit Image mode for image chaining — each output feeds the next prompt.
Use the Quick Fix button if generation gets stuck.
Increase Random Delay if you hit rate limits.
Start with 1 concurrent prompt and increase slowly.
Name character images descriptively for auto-matching: hero.png, dragon.jpg.
Download at 2K 60FPS for the highest quality output available on Seedance.

🔧 Troubleshooting ▼

❓ Extension not active

Must be on dreamina.capcut.com or jimeng.jianying.com. Navigate there and refresh.

❓ Generation errors

The tool auto-retries. Use the Quick Fix button if it stays stuck.

❓ Downloads not working

Go to Chrome Settings → Downloads and turn OFF "Ask where to save each file before downloading".

❓ Login required error

You must be logged into Dreamina / Jimeng before using the extension.

❓ Image chaining not working

Make sure Edit Image mode is enabled in Settings. Image chaining only works in Mode 5 with Edit Image active.

🚀 Getting Started ▼

Open meta.ai/media and make sure you are logged in to Meta.

Click the extension icon and pin it to the toolbar.

In the Control tab, configure: Concurrent Prompts and Prompt Delay.

Select your mode and click Run.

⚠ Important

The extension only activates on meta.ai or meta.ai/media. If the panel shows "not active", navigate to one of those URLs and refresh.

🎛️ Modes Overview ▼

Mode 1

Text-to-Video

Generate videos from text prompts. Blank line separator.

Mode 2

Image-to-Video

Animate a source image with text prompts. One image used across all prompts.

Mode 3

Text-to-Image

Generate images from descriptions with configurable aspect ratio.

Mode 4

Image-to-Image

Generate variations of a source image using text prompts.

📋 Mode Details ▼

Mode 1 — Text-to-Video

Enter prompts separated by blank lines, or upload a .txt file.
Click Run to start batch generation.

Mode 2 — Image-to-Video

Upload or drag-drop a single source image.
Enter prompts — the same image is used with each prompt.
Click Run.

Mode 3 — Text-to-Image

Enter image descriptions, one per blank-line-separated block.
Configure Aspect Ratio in Settings.
Click Run.

Mode 4 — Image-to-Image

Upload a source image.
Enter prompts describing the desired variations.
Click Run.

⚙️ Settings Reference ▼

Setting	Options	Notes
Default Mode	Mode 1–4	Mode loaded on panel open
Default Aspect Ratio	16:9 · 9:16 · 1:1	For image generation modes
Outputs per Prompt	1 – 4	Videos or images generated per prompt
Concurrent Prompts	1 – 6	Parallel prompt submissions
Prompt Delay	0 – 300 s	Wait between prompt submissions
Auto Download	On · Off	Automatically save generated media
Language	EN · VI · ZH · KO · JA · ES	UI language

💡 Tips & Best Practices ▼

Increase Prompt Delay if you hit rate limits.
Start with 1 concurrent prompt and increase slowly.
Downloads are auto-sorted into project folders for easy organization.
For Image-to-Video, use a high-quality, well-lit source image for best results.
Use English prompts for the most consistent output quality.

🔧 Troubleshooting ▼

❓ Extension not active

Must be on meta.ai or meta.ai/media. Navigate there and refresh the page.

❓ Generation errors

The tool auto-retries on failure. Wait for the retry cycle to complete.

❓ Downloads not working

Go to Chrome Settings → Downloads and turn OFF "Ask where to save each file before downloading".

❓ Login required error

You must be logged into Meta.ai before using the extension.