Meta has released Muse Wonder and Muse Wonder Full, its first frontier AI models, putting the company back among the top rivals to OpenAI and Anthropic.
The Meta Superintelligence Lab announced the pair on Monday. Wonder is larger and more capable than Meta’s earlier Muse Spark models. Meta says the new models score higher on benchmarks and handle harder jobs such as coding and long-running agent tasks.
According to TechCrunch, Meta’s testing shows Wonder and Wonder Full posting scores comparable to Anthropic’s Claude Fable 5.1 and OpenAI’s GPT-6 Astra. In Meta’s own words, the models “mark the next step on our scaling ladder and a major leap toward personal superintelligence.”
What Muse Wonder Offers
Muse Wonder is available through the Meta Model API and Muse Code. Muse Wonder Full is limited to Muse Code subscribers and will not be served through the API. Both models will also power Meta’s Muse agent.
Meta priced Muse Wonder at $4 per million input tokens and $20 per million output tokens. That puts it in the same price range as GPT-6 Sol and Claude Opus 5.5. Both Wonder models support a 1 million token context window and accept text, images, video, and audio as input.
The company stresses “active memory,” or the model’s ability to retain and reuse what a user shares over time. Meta also highlights “active focus,” meaning the model keeps working through blockers instead of stalling. Research scientist Kavya Kopparapu said Meta put “a tremendous amount of effort into making sure your model can remember what’s important to you over days, weeks, months, and beyond.”
Meta trained the models at its Prometheus data center in central Ohio. The release is the first frontier model launch since Scale AI founder Alexandr Wang restructured Meta’s AI work. Since then, Meta has shipped Muse Spark and the Muse agent but had not matched the top-tier launches from OpenAI and Anthropic. ProPakistani earlier covered Meta’s Muse Spark 1.3 update and its Muse AI agent launch.
How Wonder Compares With OpenAI And Claude
Independent scores from Artificial Analysis’s Intelligence Index (v4.3.2) place Claude Opus 5.5 at 58, Muse Wonder Full at 56, and the API-facing Muse Wonder at 52. Claude Fable 5.1 and GPT-6 Astra both sit around 53 on the same index. Claude Sonnet 5.5 also rounds to 56, so Wonder Full sits near Anthropic’s mid-tier flagship and below Opus 5.5.
On Meta’s own 14-benchmark scorecard, Muse Wonder Full leads hardest on long coding and legal agent work. Meta reports Wonder Full at 79.0 on DeepSWE 1.1 versus about 74 for Opus 5.5 and GPT-6 Astra, and 57.5 on SWE-Marathon versus 46.3 for Opus 5.5. On Harvey’s legal agent benchmark (LAB-AA), Meta puts Wonder Full at 13.1 against 4.2 for Opus 5.5.
Claude Opus 5.5 still wins several rows on Meta’s chart, including HealthBench Pro, computer use (OSWorld), Terminal-Bench, and ProgramBench, where Opus 5.5 scores 70.5 against Wonder Full’s 59.0. Meta also says Wonder tops MemoryBench against Fable, Opus, and Astra.
In short, Meta is back in the frontier race. Wonder Full looks strongest on long autonomous coding and document-heavy agent tasks in Meta’s testing, while Claude Opus 5.5 still leads the main independent index and several general benchmarks. Developers can try Muse Wonder on the public API today; Wonder Full stays behind Muse Code’s higher-tier plans.
The post Meta Enters AI Frontier With New “Wonder” Models appeared first on ProPakistani.