🔍 Read the full analysis: Muse Wonder Full Moves Into AI As Muse Wonder Picks Up A Charge on ThorstenMeyerAI.com
Get the latest gadgets delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
Meta has released Muse Wonder and Muse Wonder Full, two proprietary models evaluated by Artificial Analysis on October 12, 2026. The evaluation places Wonder Full near leading models, while the API-available Wonder scores 52 and costs $4.52 per Intelligence Index task in the analysis—far more than GPT-6.1 Sol at the same score.
Meta has released Muse Wonder and Muse Wonder Full, with an Artificial Analysis evaluation published October 12 placing the stronger model among the top performers in its benchmark set. The evaluation gives Wonder Full an Intelligence Index score of 56, while the API-available Muse Wonder scores 52; the report estimates one index task on Wonder costs $4.52, compared with $0.72 for GPT-6.1 Sol, which also scores 52.
Artificial Analysis says Muse Wonder Full improved on all 10 evaluations in its Intelligence Index v4.3 compared with Muse Spark 1.3, rising from 48 to 56. Among 27 configurations compared in the source report, only Claude Opus 5.5 at its maximum setting scored higher, at 58; Claude Sonnet 5.5 at maximum also scored 56. These are benchmark results reported by Artificial Analysis, not a guarantee of performance in every use case.
The evaluation highlights gains on Terminal-Bench 4.0, where Wonder Full scored 57%, up from 33% for Muse Spark 1.3, and GDPval-AA, where it scored 1,857 Elo. The report says Wonder Full ranked in the top three on six of the ten evaluations. Muse Wonder, the model available through Meta’s API and Muse Code, scored 52 on the index. It recorded 52% on Terminal-Bench and 68% on AutomationBench-AA, compared with 58% for Spark 1.3.
Artificial Analysis estimates a cost of $4.52 per Intelligence Index task for Muse Wonder, based on its testing and the model’s listed prices. The report compares that with $0.72 for GPT-6.1 Sol, $1.99 for Gemini 4 Argon and $5.46 for Claude Sonnet 5.5. Meta lists Muse Wonder at $4 per million input tokens and $20 per million output tokens, the same rates the report gives for Claude Opus 5.5. Wonder Full is limited to Muse Code and has no per-token price, so an outside user cannot directly calculate its cost on the same basis.
Infographic 01 · Where Meta’s new pair lands
Muse Wonder Full ties for second. Muse Wonder ties GPT-6.1 Sol.
Artificial Analysis scores 27 frontier and open-weight configurations on ten evaluations. Meta’s two new models land near the top of the list. Six weeks earlier, Muse Spark 1.3 scored 48.
Intelligence Index score, 27 configurations
Higher is better. Reasoning effort in brackets, as Artificial Analysis tested it.
Availability: Muse Wonder runs on Meta’s API and in Muse Code. Muse Wonder Full is only available to Muse Code users and has no per-token price.
Source: Artificial Analysis, Muse Wonder analysis (12 Oct 2026) and model leaderboard, read 13 Oct 2026. Intelligence Index v4.3, ten evaluations.
Powered by Thorsten Meyer AI
Strong Scores Meet High API Costs
The release puts Meta into a more competitive position on the reported benchmark comparisons. Wonder Full’s score of 56 places it close to leading configurations in the evaluation, and the report’s Coding Agents Index gives Muse Code plus Wonder Full a score of 68, tying the listed Claude Code plus Sonnet 5.5 configuration. That result applies to the combined model-and-tool setup; it does not establish how Wonder Full would score inside another coding harness.
For customers choosing an API model, the price-performance comparison may matter as much as the headline score. In Artificial Analysis’s estimate, Muse Wonder and GPT-6.1 Sol both score 52, but the task cost is $4.52 for Wonder and $0.72 for Sol. The report attributes much of Wonder’s cost to its input use: it says the model reads 7.69 million input tokens per task, with input accounting for $3.43, or 76%, of the estimated $4.52 total. Actual bills will vary with usage and pricing conditions, but the reported figures give buyers a reason to test their own workloads before switching.
The evaluation also reports a 23% hallucination rate for Wonder Full and 27% for Wonder, compared with 33% for Spark 1.3. The report describes that as improvement, while noting that some other high-scoring models have lower rates. These figures underline that a stronger benchmark score does not remove the need to check model outputs in workflows where errors carry financial or operational consequences.
As an affiliate, we earn on qualifying purchases.
Two Models, Different Access
Meta released the models about six weeks after Muse Spark 1.3, according to the source report. Both are described as proprietary models with a one-million-token context window. They accept text, images, speech or audio, and video as input, and return text. Their availability differs: Muse Wonder is offered through Meta’s API and inside Muse Code, while Wonder Full is available only to Muse Code users.
That access split limits direct comparisons for developers. API users can buy Muse Wonder at Meta’s published token rates, but the stronger Wonder Full model has no public per-token price in the source material. Artificial Analysis also reports separate scores for the model-plus-harness combinations in its Coding Agents Index: Muse Code with Wonder Full scored 68, Muse Code with Wonder scored 63, and Muse Code with Spark 1.3 scored 54. Those results measure the combined system, not the model in isolation.
The source report’s price comparisons use estimated costs for one task in Artificial Analysis’s Intelligence Index, rather than a universal charge for all prompts. It says Muse Wonder’s task used 7.69 million input tokens and about 54,000 output tokens. Because task length and token use vary, those estimates should not be read as a fixed price for a typical API request.
““The difference is the bill: one Intelligence Index task costs $4.52 on Muse Wonder and $0.72 on Sol.””
— Thorsten Meyer, ThorstenMeyerAI.com
As an affiliate, we earn on qualifying purchases.
Pricing and Real-World Performance
The source material does not include Meta’s explanation for setting Muse Wonder’s API rates at $4 per million input tokens and $20 per million output tokens, or how those rates were chosen. It also does not provide a per-token price for Wonder Full, so its cost cannot be compared directly with API models from the figures supplied.
Artificial Analysis’s index results are one evaluation, and the source report does not establish how the models perform across every customer workload. The reported task costs depend on the benchmark’s token use and pricing assumptions; they may not match the cost of a particular application. Further details about the evaluation methodology, confidence intervals for all comparisons and real-world error rates are not included in the supplied material.
The report says Wonder Full and several Claude configurations fall within each other’s confidence intervals on GDPval-AA, which makes a simple ranking on that evaluation less conclusive. It is also not yet clear whether Meta will make Wonder Full available through its API or publish a public price for it.
As an affiliate, we earn on qualifying purchases.
Watch for Wider Access and Testing
The immediate next step for prospective users is to test Muse Wonder against their own prompts, token volumes and quality requirements, then compare measured costs with alternatives. The reported $4.52 task estimate is useful as a benchmark comparison, but the source does not show what an individual customer’s bill will be.
Further information from Meta could clarify whether Wonder Full will remain exclusive to Muse Code, whether its access will expand, and whether the company will publish pricing that allows direct cost comparisons. Additional independent evaluations could also show whether the reported improvements hold across other benchmarks and practical tasks.
As an affiliate, we earn on qualifying purchases.
Key Questions
What did Meta release?
Meta released Muse Wonder and Muse Wonder Full. The source report says Wonder is available through Meta’s API and Muse Code, while Wonder Full is currently available only in Muse Code.
How did the models score?
Artificial Analysis gave Muse Wonder Full a score of 56 and Muse Wonder a score of 52 on its Intelligence Index v4.3. The evaluation was published October 12, 2026.
Why is Muse Wonder’s price being questioned?
Artificial Analysis estimates one Intelligence Index task costs $4.52 on Muse Wonder, compared with $0.72 on GPT-6.1 Sol, although both scored 52 in the report. The estimate reflects the benchmark’s token use and should not be treated as a fixed cost for every API request.
Can developers use Muse Wonder Full through Meta’s API?
Not according to the source material, which says Wonder Full is available only to Muse Code users. It does not give a public per-token price or confirm plans for API access.
Do the benchmark results mean the models will be reliable in every task?
No. The figures describe performance in Artificial Analysis evaluations. The report gives hallucination rates of 23% for Wonder Full and 27% for Wonder, so users should verify outputs for tasks where errors matter.
Source: ThorstenMeyerAI.com
Columbus Day / Indigenous Peoples' Day Picks
long weekend sales
As an affiliate, we earn on qualifying purchases.
