Precise instruction-following
Follows detailed prompts down to fine layout and typography, so dense, multi-part visuals hold together.
AvailablexAI · precise image generation and editing
Grok Imagine Image 2.0 API
Grok Imagine Image 2.0 is xAI's next-generation image model, built for real creative work rather than one-shot novelty. It follows instructions closely down to fine detail, plans typography and layout the way a designer would, and supports multi-reference editing across up to 5 input images in a single request. Access it on Muapi with the same asynchronous submit-and-poll pattern used across every model.
Grok Imagine Image 2.0 API is available on Muapi
Create a Muapi API key, submit a generation request, and poll the returned request ID for the finished image.
Grok Imagine Image 2.0 capabilities
xAI built Image 2.0 around editing as a first-class capability, not a bolt-on, alongside stronger instruction-following for generation.
Multi-reference editing
Combine up to 5 reference images in one generation — no manual compositing to merge a character, location, and props.
Targeted edits
Change only the region described in the prompt while the rest of the image is preserved.
Smart resize
Recompose the same composition into any aspect ratio, from tall banners to wide cinematic frames.
Sharp text rendering
Small text and dense typography come out legible — built for posters, infographics, and UI mockups.
Consistent style across edits
Preserves subject, style, and composition across successive generations.
Real-world performance
xAI reports Image 2.0 ranks second in the world in both text-to-image generation and image editing on the public Arena leaderboards (as of August 2026), trained for fidelity across photography, design, and illustration.
One Grok Imagine integration
Grok Imagine Image 2.0 uses Muapi's shared authentication, task lifecycle, file upload, and billing infrastructure — the same pattern already used by the Grok Imagine models.
Grok Imagine Image 2.0 API endpoints
Every Grok Imagine Image 2.0 request uses the same Muapi REST contract. Submit JSON with the x-api-key header, then poll the standard prediction endpoint.
Submit a generation task
curl -X POST https://api.muapi.ai/api/v1/grok-imagine-image-2 \
-H "x-api-key: YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{"prompt":"A high-contrast halftone portrait in fine white dots on black", "aspect_ratio":"1:1"}'Retrieve the completed result
curl https://api.muapi.ai/api/v1/predictions/REQUEST_ID/result \ -H "x-api-key: YOUR_API_KEY"
Use the returned image URL once the task status is complete. Optionally pass up to 5 reference image URLs to edit or combine existing images instead of generating from scratch.
Grok Imagine Image 2.0 API FAQ
What's new in Grok Imagine Image 2.0 compared to the original Grok Imagine?
Image 2.0 focuses on precise instruction-following, sharp typography and layout for dense visuals, and multi-reference editing that accepts up to 5 input images in one generation.
Can I edit an existing image with Grok Imagine Image 2.0?
Yes. Provide one or more reference image URLs along with a prompt describing the edit, and the model applies targeted changes while preserving the rest of the image.
How many reference images can I combine in one generation?
Up to 5 reference images can be supplied in a single request, useful for combining a subject, a location, and props into one consistent scene.
What aspect ratios are supported?
Image 2.0 supports a wide range of ratios including 1:1, 1:2, 2:1, 9:16, 16:9, 2:3, 3:2, 3:4, and 4:3, so the same composition can be recomposed for different placements.
How do I access Grok Imagine Image 2.0?
Create a Muapi API key, send it in the x-api-key header, and call the grok-imagine-image-2 endpoint.
How is Grok Imagine Image 2.0 billed?
It uses Muapi's pay-per-generation billing model, the same as the rest of the Grok Imagine lineup.