GPT Image 2.6
Create a still image from text, edit with references, or make a short animated loop.
Choose GIF for a waving mascot, a reaction, or another short movement. Describe one action and keep the scene focused. This connection creates GIFs from text only, uses Auto aspect ratio, and does not expose frame-count or duration controls. Play the returned animation before downloading to check flicker, unexpected changes, and whether the loop communicates your idea. Keep the original GIF file so the animation is preserved.
Describe the subject, composition, and lighting to create a still image from text. Choose PNG, JPEG, or WEBP based on where you plan to use the result.
Add a reference image and describe what should change and what should stay recognizable. Choose PNG, JPEG, or WEBP for the edited result. Compare the result against the originals before downloading. For reference-based workflows with other models, explore the AI Image Editor.
Flare and Sunburst share the same formats and credit price. Compare them against your own brief.
Flare is the default variant. Keep your prompt and output format fixed while you review the first result. Judge the composition, wording, or motion you actually need rather than treating a model label as a quality guarantee.
Try Sunburst when you want another interpretation of the same task. It uses the same 6-credit price per output. Compare the results before choosing; neither a variant name nor a higher quality setting guarantees that every detail will match your brief.
Pick a format first, then describe a result that suits it.
Start with PNG, JPEG, or WEBP for a still image and GIF for a short animation. If you have reference images, stay in a still-image format. GIF generation uses text only.
For a GIF, focus on one movement and keep the scene simple. For a still-image edit, upload your references and explain what should change. Select Flare or Sunburst and review the quality and complexity settings.
Play the GIF or compare a still-image edit with its references before downloading. Downloads keep the selected format. Review the credit total before generating another version; each output uses 6 credits.
From still-image edits to short loops, these comments focus on choosing an output format.
Check format restrictions, raster size, and credits before you submit.
Yes. Choose GIF in Output format and write a text prompt describing a simple loop. Still images are available as PNG, JPEG, and WEBP.
Not through this connection. GIF generation is text-only. Choose PNG, JPEG, or WEBP when working with reference images. Remove your references before switching to GIF; the workspace should not silently discard them.
No. This connection does not expose frame-count or timing controls, and GIF uses Auto aspect ratio. Describe the movement in your prompt and review the returned animation rather than expecting a specified duration or number of frames.
This AIEnhancer workspace currently offers 1K output. Inspect fine details and lettering before using any output in a final project. To explore models with other output options, visit the AI Image Generator.
Both variants and all four output formats use 6 AIEnhancer credits per output. Reference images in a still-image task are included. Two outputs use 12 credits and four use 24. A paid membership is required.