Muse Spark
Meta · Muse
Meta's multimodal reasoning model powering Meta AI, with tool use, parallel-agent thinking, voice, and Muse Image integration.
Overview
Freshness note: Model capabilities, limits, and availability can change quickly. This profile is a point-in-time snapshot last verified on July 10, 2026.
Muse Spark is the first model in Meta’s Muse family and the current reasoning layer behind Meta AI. Meta positions it as a natively multimodal reasoning model with tool use, visual chain of thought, and multi-agent orchestration. It launched in April 2026 through meta.ai and the Meta AI app and has since expanded into voice, shopping, live camera help, social context, Meta apps, and AI glasses in supported regions.
Muse Spark also now plans and shares tools with Muse Image, Meta’s July image-generation release. The pairing lets a reasoning model and image model combine search, code, media generation, and iterative planning inside one product workflow.
Capabilities
Meta’s launch claims center on four areas: multimodal perception, reasoning, health, and agentic behavior. The model is meant to understand visual information, work across tools, and support more deliberate reasoning than the ordinary “instant answer” assistant mode people associate with consumer AI chat products.
The most distinctive feature in the launch is Contemplating mode, which Meta describes as multiple agents reasoning in parallel. Meta explicitly positions that mode against extreme-reasoning offerings such as Gemini Deep Think and GPT Pro, and claims 58% on Humanity’s Last Exam and 38% on FrontierScience Research. Even if those numbers need continued independent scrutiny, they show how Meta wants this model understood: not just as a social-app assistant, but as a serious reasoning system with configurable test-time compute.
Meta also emphasizes personal use cases. The launch examples focus on visual STEM reasoning, localizing annotations in the world around you, turning prompts into playable experiences such as a Sudoku game, and supporting health-oriented explanations. Meta says it worked with more than 1,000 physicians on health-related training data to improve factuality and completeness in health reasoning.
The May product update added faster interruptible voice conversations, live camera questions, shopping across Facebook Marketplace and wider web results, and richer Reels, map, location, and community context. Rollout across WhatsApp, Instagram, Facebook, Messenger, Threads, and Meta AI glasses remains country- and device-dependent.
For media workflows, Muse Spark can jointly plan with Muse Image and combine code with image generation for outputs such as animated GIFs, websites with embedded images, and interactive visual experiences.
Technical Details
Meta describes Muse Spark as a natively multimodal reasoning model, but its current public product material does not expose a developer-style spec sheet with a token window, maximum output limit, or API rate card. The displayed contextWindow and maxOutput values therefore use 0 to mean unavailable rather than literal zero-token limits.
The more useful technical signal right now is the scaling story Meta chose to publish. The launch post highlights improvements across pretraining, reinforcement learning, and test-time reasoning. Meta says its rebuilt pretraining recipe can reach the same capability level with over an order of magnitude less compute than Llama 4 Maverick, and it also claims smoother reinforcement-learning gains plus thought compression during test-time reasoning. In plain terms, Meta is arguing that Muse Spark is a stack-level reset, not just a bigger checkpoint.
Pricing & Access
Current official availability is:
meta.ai- the Meta AI app
- staged rollout across WhatsApp, Instagram, Facebook, Messenger, and Threads
- supported Meta AI glasses in selected countries
- a private API preview for select users
Meta’s current public material does not publish token pricing or a broad self-serve API path. apiAccess: true reflects the documented private preview for selected partners, not general developer availability. Teams should treat Muse Spark as a product-first model with limited developer access.
Best Use Cases
Muse Spark is most relevant for multimodal assistant experiences where visual understanding, tool use, voice, and stronger reasoning matter together: visual troubleshooting, interactive learning flows, health explanation interfaces, shopping, and Muse Image-assisted creative work.
It is less suited to production planning when you need stable API documentation, public pricing, or long-settled enterprise availability. Today, the strongest case for evaluation is “Meta’s newest reasoning direction is now live in its own products,” not “this is already the safest default for third-party deployment.”
Comparisons
- Llama 4 Maverick (Meta): Meta’s own launch framing treats Muse Spark as a major step forward from the Llama 4-era baseline and as the first result of a rebuilt training stack.
- Muse Image (Meta): Specialized image-generation partner that can share tools and plans with Muse Spark.
- Gemini Deep Think / GPT Pro: Meta explicitly says Contemplating mode is meant to compete with the extreme-reasoning modes of those models on harder tasks, though Meta’s public developer documentation is still thinner than what those rival ecosystems expose.