AI Image Generator supports text-to-image, AI revise of uploads, caption overlays, and reusable format settings for social posts.

Choose Create or Revise with AI, pick an output format, add an optional style, generate with live progress, and reuse successful runs from My Stuff.

Go from idea to finished image in a few steps:

1. Choose a generation mode: Create or Revise with AI. Switching modes clears the current result so the next run starts clean.
2. Pick an Output format (Square, Portrait, Story, Landscape), optionally select an Image Style, or use Surprise Me to auto-fill a curated prompt and matching allowed style at no credit cost.
3. For Revise with AI, upload a base image (PNG/JPG/WEBP/GIF, up to 10MB).
4. Describe what you want: scene instructions for create, or revise instructions for AI edit.
5. Click Generate Image or Revise with AI and wait for rendering. Failed runs keep your last good image.
6. In revise mode, review the Before/After comparison.
7. Use Add text for captions/shapes, open the large preview, then Download as PNG.

• Create: Generates a new visual from text instructions at your chosen output format.
• Revise with AI: Uses your uploaded image plus edit instructions, preserving subject identity unless you ask to change it.
• Mode Reset: Switching between create and revise clears the current generated image and active error state.
• Edit Safety Rules: Upload validation enforces image MIME type and 10MB maximum size before generation.
• Add text: After generation, open the caption overlay editor to add hooks, CTAs, and shapes without another AI charge.
• Restore Limitation: Restoring a revise snapshot reloads the saved settings and format, but you still need to upload the base image again.

During rendering, you get clear progress feedback.

• Progress Percentage: Real-time generation progress with smooth updates.
• Elapsed + Estimated Remaining Time: Better timing visibility during longer renders.
• Phase Timeline: Prompt understanding, composition, detail rendering, and final quality pass.
• Mobile Flow: Mobile uses dedicated Settings, Result, and History tabs, and generation moves you toward the result view automatically.

Choose Square 1:1, Portrait 4:5, Story 9:16, or Landscape 16:9 so the model targets the right post shape. Style remains optional for look and mood.

Image generation uses the same credit wallet as other AI tools.

Successful generations are automatically saved to My Stuff with the prompt, mode, style, and source metadata needed for reuse.

• Open My Stuff from the desktop header button, the mobile History tab, or this filtered route: Image Generator My Stuff
• Build a reusable visual library for recurring campaigns without manually re-entering prompts each time.
• Restore from My Stuff to reload saved settings quickly. The result image is cleared on restore, and edit snapshots still require a fresh base-image upload.

• Specify camera language: Mention framing, lens feel, angle, and lighting for better control.
• Define constraints: Include composition rules ("centered hero", "negative space on right").
• Write for intention: Include target platform use (ad creative, cover image, carousel first slide).
• Iterate in short cycles: Keep the topic stable and adjust one variable at a time.
• Use Surprise Me strategically: Start with a curated concept when you want fast inspiration, then refine the prompt with your own brand details.
• Edit mode strategy: Keep instructions specific ("replace background", "remove logo", "add warm lighting") for cleaner transformations.
• Pair with content: For end-to-end posts, generate copy in AI Content Generator and visuals here.