ChatGPT Automation is a powerful productivity tool designed to supercharge your workflow on ChatGPT.com. Stop manually entering prompts one by one—automate the process and batch generate responses and images at scale.
- 🚀 Advanced Batch Processing: Queue dozens or hundreds of prompts and let the extension handle the submission and generation automatically.
- 🧩 Workflow (visual drag-and-drop editor): Connect prompts, images and generators on a canvas — e.g. generate images, then edit those images with more prompts automatically. Save several workflows, run one node or all of them, import/export them as files.
- 💬 Text Processing Automation: Batch generate text responses from ChatGPT. Supports multiple prompts with custom delays.
- 🧩 Ingredients to Text: Process multiple input components or images together with each prompt.
- 🖼️ Text-to-Image Batching: Create multiple images simultaneously from text descriptions.
- 🔄 Image-to-Image: Transform and enhance existing images using AI with text prompts.
- ⚙️ Professional Control Suite:
- Concurrent Prompts: Process multiple prompts simultaneously to save time.
- Smart Delays: Set custom intervals between prompts to manage rate limits effectively.
- Auto Download: Automatically save the final results to your computer.
- Auto Change File Name: Automatically rename downloaded files to keep them organized.
- Save to Folder: Specify a custom subfolder for downloaded files per project.
- 🎭 Auto-add Character Images: Automatically match and attach images to prompts based on character names in filenames.
- 📊 Real-time Queue Management: Monitor your generation progress with a visual status bar and active prompt list in the Side Panel.
- 📂 Organized File Management: Automatically sorts downloads into project-specific folders to keep your workspace clean.
- 🔧 Quick Fix: One-click tool to recover from generation errors.
- 💳 Plan Management: Sign in to check your plan and track daily prompt usage.
- 🌐 Multi-language Support: Available in English, Vietnamese, Chinese, Korean, Spanish, and Japanese.
- Visit the Chrome Web Store
- Click Add to Chrome.
-
Navigate to ChatGPT
- Open chatgpt.com
- The extension only works on ChatGPT pages.
-
Open the Extension
- Click the extension icon in Chrome toolbar. Pin it for easier access!
-
Configure Batch Settings
- In the Control tab, you can set:
- Concurrent Prompts: How many prompts to run at the same time.
- Prompt Delay: Wait time between each prompt submission.
- Save to Folder: Subfolder name for downloaded files.
- Auto Change File Name: Toggle automatic file renaming.
- In the Control tab, you can set:
-
Select a Mode
- Choose from the available modes: Text Processing, Ingredients to Text, Text to Image, or Image to Image.
- Select Text Processing mode.
- Enter prompts into the input box (separate each prompt with a blank line).
- Alternatively, click or upload .txt file to import a list of prompts from a
.txtfile. - Click Run to start the batch.
Example Prompts:
Write a professional email to a client about a project delay.
Keep it concise and apologetic.
Summarize the key benefits of remote work in 5 bullet points.
Write a product description for a noise-cancelling headphone.
Target audience: remote workers.
- Select Ingredients to Text mode.
- Upload the images or component files you want to include.
- Enter prompts (separate with blank lines). Images will be processed with each prompt.
- Optionally enable Auto-add character images to automatically attach images whose filenames match character names mentioned in your prompts.
- In Settings, set Max Input Images per Prompt (1–3) to control how many images are used per prompt.
- Click Run.
- Select Text to Image mode.
- Enter detailed descriptions for your images (separate each prompt with a blank line).
- Configure the desired Image Model in the Settings tab.
- Click Run.
Example Prompts:
A serene mountain landscape at sunrise with golden light reflecting on a lake.
A futuristic city skyline at night with neon lights and flying cars.
A cozy coffee shop interior with warm lighting and bookshelves on the walls.
- Select Image to Image mode.
- Upload the source images you want to transform.
- Enter prompts (separate with blank lines).
- Optionally enable Auto-add character images to automatically attach images based on filename matching.
- In Settings, set Max Input Images per Prompt (1–10).
- Click Run.
Workflow is a visual drag-and-drop editor for flows with several steps — for example: generate a few images, then edit those images with more prompts. It opens in its own window and runs on your open chatgpt.com tab.
- Click Workflow in the Control tab (bottom row).
- Already typed prompts or uploaded images in the side panel? Hover Workflow and click Convert to workflow: your prompts, each prompt's mode and your images become nodes in the editor, ready to run.
| Area | What it holds |
|---|---|
| Left | Nodes (click or drag one onto the canvas) and Your workflows (all saved workflows) |
| Top bar | The canvas tools: Undo/Redo, Auto arrange, fit view, Example, clear. On the right: the Details button, Shortcuts and the chatgpt.com tab status |
| Canvas | Your nodes. Top-left: Run all (and Stop while running) and Enable background mode |
The Details button shows what needs attention: Issues (n) in red/yellow when something blocks a run, Running 3/8 while generating. Click it to open a panel with the issues (click one to jump to the node), live progress, the run plan and the settings it uses.
| Node | What it does |
|---|---|
| Enter prompt | One or more prompts, separated by a blank line |
| Upload image | Your images (drop files on it). Hover an image: 🔍 to view it larger, ✕ to remove it, the grip in the corner to drag it to another position. The order (or the sort menu) decides which prompt gets which image |
| Generate Image | Text to Image, or Image to Image when images are connected. Options: Image Mode per Prompt, Max Input Images per Prompt, Auto-add character images |
Generate nodes are named automatically from their first prompt (image_… / video_…). Each prompt row shows the images it will receive, so you can check before running. Their preview takes the aspect ratio from the settings (a 9:16 node is narrower and taller).
Drag from the round handle on the right of a node and drop it anywhere on the other node — the right input is picked for you. Nodes that accept the connection light up while you drag.
| From | To | Meaning |
|---|---|---|
| Enter prompt | Generate Image | The prompts to generate |
| Upload image | Generate Image | Reference images, start frames or components |
| Generate Image | Generate Image | The generated images become that node's input (it runs after the images are ready) |
- Run all (top-left, or
Ctrl/⌘ + Enter) runs the whole workflow in the right order: nodes waiting for generated images start automatically once those images exist. - If Run all is disabled, the top bar shows Issues (n): click it to see what to fix.
- Each Generate node also has its own Run button to run only that node. It is disabled until the nodes it depends on have finished (hover to see why).
- Stop cancels what is still running.
- While running, the connections into the node that is generating light up and flow, so you can see where the workflow is.
⚠️ Chrome pauses chatgpt.com when its tab isn't visible (for example when the workflow window covers it full screen). Click Enable background mode (under Run all in the workflow, or in the side panel), then pick the chatgpt.com tab in Chrome's dialog. This shares the chatgpt.com tab (nothing is recorded or sent anywhere) so it keeps generating behind other windows. The green Running in background badge shows it's on; click ✕ to stop it.
Results appear inside each Generate node. Hover a result: 🔍 opens it large, ✕ removes it (the eraser clears all results of the node). Files are still downloaded as usual.
The next node uses the first result of each prompt. To choose which one, drag a result by the grip in its top-left corner onto another result to swap them.
Under Your workflows (left): New, Import, and for each workflow the ⋯ menu — Rename (or double-click the name), Duplicate, Export, Delete. Everything is saved automatically.
- Export downloads a
.jsonfile. It starts with//comment lines that describe every node, property and connection, so you can give the file to an AI assistant and ask it to write new workflows. The comment lines are removed on import. - Import a file with the button, or simply drag the
.jsonfile onto the canvas.
Click Shortcuts in the top bar (or press ?) to see them all.
| Action | Keys |
|---|---|
| Undo / Redo | Ctrl/⌘ + Z / Ctrl/⌘ + Shift + Z |
| Copy / Cut / Paste nodes (also into another workflow) | Ctrl/⌘ + C / X / V |
| Duplicate selection | Ctrl/⌘ + D |
| Select all / Add to selection / Box select | Ctrl/⌘ + A / Ctrl/⌘ + click / Shift + drag |
| Auto arrange | Shift + A |
| Delete selected | Delete |
| Run all / Run in background | Ctrl/⌘ + Enter / Ctrl/⌘ + Shift + Enter |
Access the Setting tab to customize your experience:
| Setting | Description |
|---|---|
| Default Mode | Set which mode opens by default. |
| Outputs per Prompt (Text) | Set how many text responses (1–4) to generate per prompt. |
| Outputs per Prompt (Image) | Set how many images (1–50) to generate per prompt. |
| Concurrent Prompts | Number of prompts to process simultaneously (1–6). |
| Random Delay | Random wait time before handling the next prompt. |
| Text Model | Select the text generation model to use. |
| Image Model | Select the AI model for text-to-image generation. |
| Default Prompt Mode Option | Default mode for prompts: New Chat or Concat (combine with next). Last prompt always uses New Chat. |
| Default Image Mode Option | Default input mode for image prompts: New Image or Last Image (chain from previous output). |
| Max Input Images per Prompt (Ingredients) | For Ingredients to Text mode: 1–3 images per prompt. |
| Max Input Images per Prompt (Image to Image) | For Image to Image mode: 1–10 images per prompt. |
| Max Retries on Failure | How many times to retry if generation fails (1–20). |
| Auto Download (Text) | Download option for text generation: No Download or custom. |
| Auto Download Quality (Image) | No Download or 1k quality. |
| Language | Switch between English, Tiếng Việt, 中文, 한국어, Español, 日本語. |
- Wait Times: If you encounter rate limits, increase the Random Delay in Settings.
- Concurrent Runs: Start with 1 concurrent prompt and increase slowly to see what your account supports.
- Prompting: Be specific! Detailed prompts lead to better AI generations. Separate multiple prompts with a clear blank line.
- File Organization: Use the Save to Folder field to keep project downloads in named subfolders.
- Character Images: Name your image files after characters (e.g.,
hero.png,villain.jpg) and enable Auto-add character images to have them matched automatically. - Prompt Chaining (Concat): Use the Concat mode option to chain prompts together — each prompt is combined with the next in a single ChatGPT conversation.
- Image Chaining (Last Image): Use the Last Image mode option for Image to Image — each prompt's output becomes the next prompt's input.
- Quick Fix: If a generation gets stuck or errors, use the Fix Error button in the Control tab.
| Issue | Solution |
|---|---|
| Extension not active | Ensure you are on chatgpt.com. Refresh the page (Ctrl+R or F5) if needed. |
| Generation Errors | ChatGPT might be busy. The extension will automatically retry based on your Max Retries setting. Use Fix Error to manually recover. |
| Downloads not working | Ensure "Ask where to save each file before downloading" is OFF in Chrome Settings. |
| Login Required | Make sure you are logged into your ChatGPT account. |
| Connection Error | Refresh the ChatGPT page (Ctrl+R, F5) and try again. If the problem persists, try reinstalling the extension. |
| Workflow: generation stays at "Generating" forever | Chrome paused the hidden chatgpt.com tab. Turn on Enable background mode (or Run in background), or keep chatgpt.com visible. |
| Workflow: a node's Run button is disabled | Hover it: run the node it depends on first, or fix the issue shown (e.g. no prompt connected). |
| Workflow: Run all is disabled | Click Issues (n) in the top bar to see what to fix; click an issue to jump to its node. |
| Workflow: "No chatgpt.com tab found" | Open chatgpt.com in a tab (the green dot in the top bar shows it's connected). |
- Local Processing: All automation logic runs locally in your browser.
- No Data Collection: We do not store or collect your prompts, images, or account data.
- Secure Storage: Settings are saved only in your browser's local storage and synced across tabs.
- Author: Trường Nguyễn
- Website: kylenguyen.me
- Feedback: Use the "Report a bug" link in the extension.
Current version: 1.0.2
Copyright © 2026 Trường Nguyễn. All Rights Reserved.
This software is proprietary. Unauthorized copying or distribution is prohibited.
Made with ❤️ by Trường Nguyễn