Meta Unveils Muse Wonder AI Models, Claims Parity With OpenAI and Anthropic
Meta released AI models Muse Wonder and Muse Wonder Full, which it says match OpenAI and Anthropic; Artificial Analysis scored Full at 56 against 53 for GPT-6 Astra.
Why it matters
The release puts Meta into direct competition with OpenAI and Anthropic and sets a per-token price below both rivals' flagship models. No crypto asset is involved, so the significance for crypto markets is limited; the parity claim itself is unsettled, since one third-party evaluation ranked standard Wonder below two of the models Meta says it matches and Abacus.AI's CEO disputed the comparison publicly.
Where the sources disagree
Whether Muse Wonder rivals leading OpenAI and Anthropic models
Meta claims Muse Wonder and Muse Wonder Full match models from OpenAI and Anthropic
CoinGape, Cointelegraph
Artificial Analysis: Wonder Full 56, Wonder 52, versus Claude Fable 5.1 and GPT-6 Astra 53 each and Claude Opus 5.5 58
Cointelegraph
Bindu Reddy: Muse Wonder was nowhere near Fable or Astra level
Cointelegraph
Meta says the models match up to OpenAI and Anthropic, while Artificial Analysis scores put standard Wonder below GPT-6 Astra and Claude Fable 5.1 and Abacus.AI's CEO rejected the comparison outright.
Key facts
Meta released two AI models, Muse Wonder and Muse Wonder Full.
Reported by CoinGape
Meta called the models its most capable to date and a step toward personal superintelligence.
Reported by CoinGape
Artificial Analysis scored Muse Wonder Full at 56 and Muse Wonder at 52 on its Intelligence Index, both at maximum reasoning effort.
56Disputed figureReported by Cointelegraph
On that same index, Claude Fable 5.1 and GPT-6 Astra each scored 53 and Claude Opus 5.5 scored 58.
53Disputed figureReported by Cointelegraph
Bindu Reddy, CEO of Abacus.AI, wrote on X that Muse Wonder was nowhere near Fable or Astra level and said detailed benchmark results would follow.
Disputed figureReported by Cointelegraph
Wonder Full is available only to Muse Code subscribers on the $50 a month power usage or $200 a month maximum usage tiers.
50 USDReported by Cointelegraph
Meta said the models add two capabilities it calls Active Memory, which carries relevant information across conversations, and Active Focus, which helps the agent sustain work on difficult tasks.
Reported by Cointelegraph
What happened
Meta unveiled two AI models, Muse Wonder and Muse Wonder Full, calling them its most capable models to date and a step toward what it terms personal superintelligence. The launch follows Muse Spark, the first model from Meta's superintelligence team released earlier this year, and Zuckerberg has said Meta has caught up with OpenAI and Anthropic and will keep spending to advance its AI work. Cointelegraph reports that Artificial Analysis scored Wonder Full at 56 and the standard Wonder at 52 on its Intelligence Index at maximum reasoning effort, which puts Full above Claude Fable 5.1 and GPT-6 Astra at 53 each but below Claude Opus 5.5 at 58. Bindu Reddy, CEO of Abacus.AI, rejected the comparison on X, writing that Muse Wonder was nowhere near Fable or Astra level, and said detailed benchmark results from his firm would follow. Standard Wonder charges $4 per million input tokens against $10 for GPT-6 Astra and $50 for Claude Fable 5.1, while Wonder Full is reserved for Muse Code subscribers on the $50 and $200 monthly tiers. CoinGape, citing Meta's blog post, reported the company as claiming the models match up to models from OpenAI and Anthropic, and described Meta's new models as carrying the Active Memory and Active Focus capabilities.
How the story developed
- First report
First report by CoinGape
Meta Releases Muse Wonder, Claims Frontier Capabilities to Rival OpenAI and Anthropic
- Update
CoinGape reports Meta's release of Muse Wonder and Muse Wonder Full
First report in the set, also noting the release follows Muse Spark earlier in the year.
- Official statement
Meta blog post presents the models as its most capable to date
The post said the two models mark the next step on Meta's scaling ladder and a major leap toward personal superintelligence.
- Independent corroboration
Independently corroborated by Cointelegraph
Meta claims new Muse Wonder AI models rival OpenAI, Anthropic
- Update
Cointelegraph corroborates the release with benchmark scores and pricing
Reports Artificial Analysis scores, per-token prices and subscription-only access to Wonder Full.
- Update
Abacus.AI CEO Bindu Reddy challenges the performance claims on X
Reddy said Muse Wonder was nowhere near Fable or Astra level and that detailed benchmark results would follow.
How coverage built up
Independent sources over time, counted the way the consensus panel counts them: a republication of a wire story does not move the line.
- 12 Oct 2026, 18:10 UTC: CoinGape — 1 independent source, 1 reports
- 13 Oct 2026, 01:54 UTC: Cointelegraph — 2 independent sources, 2 reports
Affected entities
Original sources
| Publisher | Report | Role | Published |
|---|---|---|---|
| CoinGape Crypto media | Meta Releases Muse Wonder, Claims Frontier Capabilities to Rival OpenAI and Anthropic | Original | |
| Cointelegraph Crypto media | Meta claims new Muse Wonder AI models rival OpenAI, Anthropic | Independent |