Muse Image

By Meta Image creation

Generate and edit images with Meta's Muse Image. Agentic grounding searches the web for real brands, landmarks and current facts before it renders, at a flat 1 token per image.

Description

**Muse Image** is Meta's agentic image generation and editing model (`muse-image-1.0`). One model handles both generation and editing, and it grounds itself against the real world before it renders.

**Capabilities**
- **Agentic grounding at no extra cost** — runs its own web search and image search before rendering, so real landmarks, products, brands and current facts come out accurate. Included in the per-image price.
- **Precise editing** — send an existing image with an instruction and it changes only what you ask, keeping the rest intact.
- **Multi-image composition** — pass up to 10 reference images to fuse subjects and settings into a single coherent scene.
- **Infographics and structured graphics** — computes and arranges charts, tables and layouts via internal code execution before rendering.
- **Flat, predictable pricing** — 1 token per image, whatever the size, format or reasoning strength.
- **Speed control** — single-pass drafts or multi-pass refinement, at the same price.

**Use Cases**
- Infographics, charts and data visualisations that need correct, current numbers
- Marketing and product imagery featuring real products, packaging or brand logos
- Travel and location imagery of real landmarks with accurate likeness
- Posters, social graphics and thumbnails with precise typography
- Iterative concept exploration where low cost per image matters
- Compositing a subject from one photo into the setting of another

Available tools

Edit Image with Muse Image

Edits one or more existing images with a text instruction, changing only what you ask while preserving the rest. Passing multiple images composes them into a single scene.

Generate Image with Muse Image

Generates images from a text prompt using Meta Muse Image. Automatically searches the web and image sources for real brands, landmarks, products and current facts before rendering, making it ideal for infographics, data graphics and factually accurate imagery.