Meta · 2026-07-07 · major
Meta Muse Image — MSL's first image model plans, calls tools, and self-refines
Meta Superintelligence Labs' first image model launches today in the Meta AI app, Instagram Stories US, and WhatsApp, ranking #2 on LMArena for text-to-image and single- and multi-image editing.
Meta Superintelligence Labs' first image model plans, calls tools, and self-refines like a reasoning LLM.
Quick facts
| Maker | Meta Superintelligence Labs (MSL) |
|---|---|
| Sibling in preview | Muse Video (text-to-video) |
| Availability | Meta AI app, meta.ai, Instagram Stories (US), WhatsApp (select countries) |
| Access | Free for everyday use; higher volume via Meta subscription plans |
| License | Proprietary, no open weights |
| Arena rank | #2 text-to-image, single- and multi-image edit (July 5, 2026) |
| Safety | Content Seal invisible watermark; detector at meta.ai/identification |
What is it?
Muse Image is the first image-generation model from Meta Superintelligence Labs, and it ships today inside the Meta AI app, meta.ai, Instagram Stories in the US, and WhatsApp Meta AI chats in select countries. Its video sibling, Muse Video, launched in early preview at the same time.
How does it work?
Instead of jumping straight from prompt to pixels, Muse Image acts as an agent — it plans, calls search and code tools for factual grounding, and self-refines its own output. Meta says this self-refinement behavior emerged during reinforcement-learning training, and that quality keeps improving as more test-time compute is spent, the same lever that powers reasoning LLMs.
Why does it matter?
Meta now has a top-tier consumer image model of its own, so Instagram, WhatsApp, and Ads can generate and edit images without paying an outside API. LMArena ranks Muse Image #2 for text-to-image, single-image editing, and multi-image editing as of July 5, and every generated image carries the Content Seal invisible watermark for provenance.
Who is it for?
Meta AI users, Instagram creators, and advertisers using Advantage+
Frequently asked questions
- Is Muse Image free to use?
- Muse Image is free for everyday creation inside the Meta AI app, meta.ai, Instagram Stories in the US, and WhatsApp in select countries. Meta says higher-volume creators can unlock extra quota through Meta AI subscription plans, but has not published the tiers or prices.
- Is Muse Image open weight?
- No. Muse Image is a proprietary Meta Superintelligence Labs model and only runs inside Meta's own apps and surfaces. Meta has not published weights, model card, parameter count, or a public API, unlike its earlier Llama releases.
- How does Muse Image compare to Muse Spark?
- Muse Spark is MSL's reasoning model; Muse Image is its image-generation model. They work together: Muse Image can call Muse Spark for multi-step planning behind a complex prompt, letting the image model reason through composition before drawing.
- What is Muse Video?
- Muse Video is Muse Image's video sibling, launched today in early preview for creators and Meta AI. Meta ranks Muse Video #3 on LMArena for text-to-video as of July 5, 2026, and says a broader rollout will follow.
- Where does Muse Image show up first?
- Muse Image is rolling out inside the Meta AI app and meta.ai, powering 30+ new Instagram Stories effects in the US, and enabling image generation in WhatsApp Meta AI chats in select countries. Meta says Facebook and Messenger follow, plus Advantage+ ad creative for advertisers.
Try it
Open the Meta AI app or meta.ai and prompt an image