🔍 Read the full analysis: Muse Wonder Full Joins The Frontier. Muse Wonder Charges Like It Did Too. 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 released Muse Wonder and Muse Wonder Full, two proprietary models evaluated by Artificial Analysis on October 12, 2026. Wonder Full scored 56 on the Intelligence Index, while the API-available Wonder scored 52; Artificial Analysis estimated a Wonder task at $4.52, compared with $0.72 for GPT-6.1 Sol at the same score.
Meta has released Muse Wonder and Muse Wonder Full, two proprietary AI models that received strong scores in an Artificial Analysis evaluation published October 12, 2026. Wonder Full scored 56 on the Intelligence Index v4.3, while the version available through Meta’s API scored 52; the evaluation estimated that a Wonder task costs $4.52, far more than a similarly rated GPT-6.1 Sol task.
Artificial Analysis reported that Muse Wonder Full scored 56, eight points above Muse Spark 1.3’s 48, and improved on all ten evaluations in the index. Among the 27 configurations compared in the source report, only Claude Opus 5.5 at maximum settings scored higher, with 58; Claude Sonnet 5.5 at maximum settings also scored 56. Wonder Full placed in the top three on six of the ten evaluations.
The two models are not equally accessible. Muse Wonder is available through Meta’s API and in Muse Code, Meta’s coding agent, at list prices of $4 per million input tokens and $20 per million output tokens. Muse Wonder Full is available only to Muse Code users and has no per-token price, leaving its cost per task unavailable for comparison. Both models have a reported one-million-token context window, accept text, images, audio and video as input, and return text.
For a single Intelligence Index task, Artificial Analysis estimated $4.52 for Muse Wonder. Its comparison put GPT-6.1 Sol at $0.72 for a score of 52, Gemini 4 Argon at $1.99 for 53, and Claude Sonnet 5.5 at $5.46 for 56. These are evaluation-specific estimates, not a universal price for every use; actual bills depend on workloads and token use.
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, Uneven Access
The release gives Meta a strong result in independent model evaluations, particularly through Wonder Full, but the highest-scoring model is not available as a priced API offering. Developers comparing models by score, availability and cost therefore cannot treat Wonder Full and Wonder as interchangeable options.
The price comparison raises a practical question for API customers: whether Wonder’s capabilities justify its estimated cost on their own tasks. In Artificial Analysis’s index-task comparison, it matched GPT-6.1 Sol’s score of 52 but cost $4.52 to Sol’s $0.72. That finding is specific to the evaluation setup, but it gives buyers a concrete reason to test their own token usage and task performance before choosing a model.
Benchmark gains also do not remove the need to check outputs. The source report put Wonder Full’s hallucination rate at 23% and Wonder’s at 27%, compared with 33% for Spark 1.3. Those reported rates indicate improvement over Meta’s predecessor, not that errors are rare or that results are reliable for every use.
As an affiliate, we earn on qualifying purchases.
From Spark to Two Models
The release follows Muse Spark 1.3 by six weeks. In the Artificial Analysis comparison, Wonder Full’s Intelligence Index score rose from 48 to 56, while the API model Wonder reached 52. The report also described gains in individual tasks: Wonder Full scored 57% on Terminal-Bench 4.0, up from Spark 1.3’s 33%; Wonder scored 52% on that benchmark.
Artificial Analysis also evaluated coding-agent combinations, which include both a model and its software harness. It scored Muse Code with Wonder Full at 68, equal to Claude Code with Sonnet 5.5. Muse Code with Wonder scored 63, and with Spark 1.3 scored 54. The report cautioned that this tests the model-and-harness combination; it does not establish how Wonder Full would perform inside Claude Code.
The benchmark results are independent evaluations reported by Artificial Analysis, not proof that every user will see the same outcomes. The source material does not provide enough methodological detail to independently assess every estimate, including the per-task cost figures, so those numbers should be read as results under the evaluator’s stated comparison rather than fixed costs for all deployments.
“The bill is in the input, not the output.”
— Thorsten Meyer, ThorstenMeyerAI.com
AI model performance evaluation tools
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
API Costs and Real-World Results
The cost of using Wonder Full remains unknown because Meta does not list a per-token price for the Muse Code-only model. The source report also does not give a comparable per-task cost for Wonder Full, so its benchmark lead cannot be directly weighed against its expense.
Artificial Analysis’s reported task costs reflect its index evaluation and should not be treated as fixed charges for other workloads. The source says Wonder consumed 7.69 million input tokens per index task and produced about 54,000 output tokens; it also attributes $3.43, or 76% of the estimated $4.52 total, to input. The supplied source text ends mid-sentence while discussing the remaining cost breakdown, so no further component is confirmed here.
It is also unclear from the available material how the models perform across customers’ own applications, what access limits apply, or whether Meta plans to make Wonder Full available through its API. Benchmark scores and estimated hallucination rates do not settle reliability for a particular high-stakes workflow.
text and multimedia input AI models
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
What Buyers Should Watch
Developers considering Muse Wonder can compare its API price with alternatives using their own prompts, context sizes and output requirements. Token consumption is a central variable: the reported index task used substantially more input tokens for Wonder than several competing models, so a benchmark score alone may not predict a customer’s bill.
For Muse Code users, the next practical test is whether Wonder Full’s reported coding-agent performance translates to their software tasks. The published score compares complete agent configurations, so customers should evaluate the model together with the available tools and workflow rather than infer that the model will rank the same in a different setup.
Further details from Meta on Wonder Full’s pricing, broader availability and deployment terms would clarify how customers can compare it with the API model. No timetable for such changes is included in the source material.
As an affiliate, we earn on qualifying purchases.
Key Questions
What did Meta release?
Meta released Muse Wonder and Muse Wonder Full, proprietary models that accept text, images, audio and video as input and return text. Both are reported to have a one-million-token context window.
How did Muse Wonder Full score?
Artificial Analysis gave Wonder Full a 56 on its Intelligence Index v4.3, eight points above Muse Spark 1.3. In the reported comparison, Claude Opus 5.5 scored 58 and Claude Sonnet 5.5 also scored 56.
Can developers access Muse Wonder Full through an API?
The source report says Wonder Full is available only to Muse Code users. Muse Wonder is available through Meta’s API and inside Muse Code.
Why is Muse Wonder’s cost drawing attention?
Artificial Analysis estimated $4.52 per Intelligence Index task for Wonder, compared with $0.72 for GPT-6.1 Sol, which received the same score of 52. The figures are tied to that evaluation and are not guaranteed costs for other workloads.
What remains unknown about the release?
The available source does not state Wonder Full’s per-token or per-task cost, whether it will become available through the API, or how the models perform across customers’ individual applications.
Source: ThorstenMeyerAI.com
Columbus Day / Indigenous Peoples' Day Picks
long weekend sales
As an affiliate, we earn on qualifying purchases.
