Muse Wonder Full
- Provider
- Meta
- Status
- Current
- Context
- 1,000,000 tok
Muse Wonder Full is the stronger of the two frontier models Meta released on 12 October 2026 (Meta). It is available only to Muse Code subscribers: it is not on the Meta Model API and has no price per token. The model developers can buy by the token is Muse Wonder.
Artificial Analysis scored it 56 on the Intelligence Index v4.3.2 at max effort, level with Claude Sonnet 5.5 and two points below Claude Opus 5.5 (58). Meta published no SWE-bench figure, no parameter count and no knowledge cutoff.
Quick specs
| Muse Wonder Full | |
|---|---|
| Released | 12 October 2026 |
| Access | Muse Code subscribers only |
| Meta Model API | Not available |
| Price per token | None published |
| Muse Code plans that include it | Power Usage, $50 a month; Maximum Usage, $200 a month |
| Context window | 1,000,000 tokens |
| Max output | Not published |
| Knowledge cutoff | Not published |
| AA Intelligence Index | 56 (max) |
| Cost per AA index task | Not measurable without an API |
Muse Code plan prices are from VentureBeat, which reports the Power Usage plan at 20 times and the Maximum Usage plan at 80 times the standard allowance.
How it differs from Muse Wonder
| Muse Wonder Full | Muse Wonder | |
|---|---|---|
| AA Intelligence Index (max) | 56 | 52 |
| Meta Model API | No | Yes, $4 / $20 |
| Muse Code | Subscribers only | Yes |
| Output tokens to run the AA index | 140 million | 130 million |
Meta has not said how the two differ technically. Its launch post says both “build on larger, stronger base models” than Muse Spark 1.3, and that Muse Wonder Full “performs even better on unfamiliar tasks”. Artificial Analysis lists them as separate models and measured both at max effort.
Benchmark performance
Independent, from Artificial Analysis:
| Model | Intelligence Index v4.3.2 |
|---|---|
| Claude Opus 5.5 | 58 |
| Muse Wonder Full | 56 |
| Claude Sonnet 5.5 | 56 |
| GPT-6 Astra | 53 |
| Claude Fable 5.1 | 53 |
| Gemini 4 Argon | 53 |
| Muse Wonder | 52 |
Artificial Analysis shows no price, output speed or cost per task for Muse Wonder Full, because there is no public API to measure.
Published by Meta, as reported by VentureBeat, which notes the table mixes internal and third-party evaluations run under varying conditions:
| Benchmark | Muse Wonder Full | Muse Wonder | GPT-6 Astra | Claude Opus 5.5 |
|---|---|---|---|---|
| DeepSWE 1.1 (software engineering) | 79.0 | 78.8 | 74.1 | 74.2 |
| SWE-Marathon (long-running coding) | 57.5 | 53.1 | 42.5 | 46.3 |
| Harvey LAB-AA (legal) | 13.1 | 10.0 | 8.6 | 4.2 |
| AutomationBench-AA (workflows) | 45.4 | 39.1 | 41.6 | 41.2 |
| JobBench (professional) | 75.3 | 70.5 | 66.1 | 75.4 |
| Terminal-Bench 4.0-AA | 57.1 | 52.0 | 59.1 | 59.6 |
| HealthBench Pro | 62.3 | 60.5 | 65.3 | 65.6 |
On Meta’s own table Muse Wonder Full leads GPT-6 Astra and Claude Opus 5.5 on DeepSWE 1.1, SWE-Marathon, the Harvey legal benchmark and AutomationBench, and trails both on Terminal-Bench 4.0 and HealthBench Pro.
Data use
Muse Code subscriptions default to Contributor mode, which lets Meta use submitted prompts, code, files and outputs to improve future models. Subscribers can switch it off, which applies to future submissions only (VentureBeat).
Known limitations
No API. Muse Wonder Full cannot be called from your own software or bought by the token.
No SWE-bench figure. Meta published no SWE-bench Pro or Verified result.
Specs withheld. Parameter count, knowledge cutoff, maximum output length and how the model differs from Muse Wonder are not published.
No open weights. Neither Wonder model has downloadable weights.
Safety report pending. Meta said a full report with additional safety evaluations “will be available soon”; it was not published at launch.
Frequently asked questions
How do I get Muse Wonder Full?
Through a Muse Code subscription. VentureBeat reports it is included in the Power Usage plan at $50 a month and the Maximum Usage plan at $200 a month. It is not on the Meta Model API.
Does Muse Wonder Full have an API?
No. Only Muse Wonder is on the Meta Model API, at $4 per million input tokens and $20 per million output tokens.
How good is Muse Wonder Full?
Artificial Analysis scores it 56 on its Intelligence Index, level with Claude Sonnet 5.5 and two points below Claude Opus 5.5, the highest-scoring model on that index.
What is the difference between Muse Wonder Full and Muse Wonder?
Muse Wonder Full scores four points higher on the Artificial Analysis index (56 to 52) and is limited to Muse Code subscribers. Muse Wonder is sold by the token on the Meta Model API. Meta has not published how the two differ technically.
Is Muse Wonder Full open source?
No. It is proprietary, with no downloadable weights.