Muse Wonder
Meta's proprietary frontier creative model, built for richly multimodal generation, art direction, and brand storytelling rather than raw open-weight efficiency.
At a glance
Released
Jun 18, 2026
Provider
Meta
Context window
400,000 tokens
Max output
32,000 tokens
Modalities
Text + image + audio → text + image
Knowledge cutoff
March 2026
API model ID
muse-wonderLicense
Proprietary (API-only)
Pricing
Standard
Input$4 / 1M
Output$20 / 1M
* Rates are from a third-party aggregator — verify before relying on them. A typical 1,500-token prompt with a 650-token answer costs ≈ $0.019 on the standard tier. Run your own prompt through the calculator →
What's new vs the previous model
Compared withMuse 1 (October 2025)
- Sharper creative control and stronger image generation
- Added audio understanding and longer creative context
- Premium pricing well above the Llama line
Best use cases
Creative campaigns
Generates on-brand copy and imagery from a single creative brief.
Storyboarding
Turns loose concepts into illustrated sequences and scripts.
Multimodal content studios
Powers tools that blend text, image, and audio inputs into finished assets.
Strengths & limitations
Strengths
- Top-tier creative quality
- Native image output
- Multimodal input breadth
Limitations
- Expensive
- Less suited to pure analytical reasoning
How to test it
- No-code: open meta.ai and pick Muse Wonder from the model switcher — no API key needed.
- API: Muse Wonder has no public API — it is available only inside Meta's apps.