Muse Wonder Full: score and benchmark evidence
Historical release · · Meta
138.5 points · Rank 5 among 38 ranked models · Stability range 128.5–147.9 points.
Rank follows the point estimate; it does not establish statistically significant superiority. The scale is anchored at mean 100, SD 15 in the frozen calibration cohort, not human IQ or a percentage.
One score summarizes demonstrated capability across published reasoning settings and seven equally weighted domains. Related benchmark variants share a family budget, and influence from one evaluator is capped.
Release tli-2026-v1.0-2026-10-28-refresh-45. Methodology and limitations for this release.
This page preserves historical evidence. See the current model record.
Capability domains
Domain scores use calibration-cohort standard-deviation units. Unmeasured domains are shown as unavailable, not zero.
| Domain | Score (SD units) | Observed benchmarks |
|---|---|---|
| Knowledge & reasoning | 1.00 | 5 |
| Coding & software engineering | 1.22 | 5 |
| Agentic tool & computer use | 1.16 | 2 |
| Professional & real-world work | 1.20 | 11 |
| Multimodal & vision | Unavailable | 0 |
| Cybersecurity | Unavailable | 0 |
| Preference & communication | Unavailable | 0 |
Published benchmark evidence
23 results contribute to this score, from 24 collected results and 23 contributing benchmark families. Counts are not independent sample sizes.
Results, source citations, review decisions and compatible reasoning modes are stored in a versioned relational database. Each publication uses a sealed, checksum-verified export; CSV files are optional exports, not editable scoring inputs. Previous revisions remain available for audit. Collected means a numeric result exists in the database. Used means its source, exact model checkpoint and benchmark protocol have been reviewed and its benchmark has a frozen calibration supported by multiple providers. For each benchmark we select the highest published aggregate across comparable reasoning modes, not necessarily maximum effort, and count it once. More tested settings provide more chances to record a high score; this selection advantage is not corrected by the conditional interval. Developer-reported and sole/unspecified settings remain eligible and disclosed. Unresolved does not mean false or useless: conflicting metrics, unverified citations and unmatched versions stay collected until reconciled.
The score stability range is a 95% bootstrap score interval conditional on frozen calibration and comparable evidence availability. It resamples benchmark families and evaluators and adds a frozen missing-evidence correction. It excludes selection across reasoning settings, selective publication, calibration changes and new domains. It is not a 95% guarantee about true capability or where a future score will land. Sparse or conflicting evidence can widen the range.
Reporting sensitivity: ±7.4 points. Sensitivity to selectively published results. This stress range is not a confidence interval or a probability statement.
AA-Briefcase v1.1 (Elo): 1729.32 native-points
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Muse Wonder Full (max); Artificial Analysis independent evaluation, 2026-10-15; fallback not reported
Metric: published native-points
Original source · Retrieved 2026-10-15
Exact model and effort in the evaluator chart payload.
GDPval-AA v2.1 (Elo): 1857.09 native-points
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Muse Wonder Full (max); Artificial Analysis independent evaluation, 2026-10-15; fallback not reported
Metric: published native-points
Original source · Retrieved 2026-10-15
Exact model and effort in the evaluator chart payload.
AutomationBench-AA: 71.3307497322%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Muse Wonder Full (max); Artificial Analysis independent evaluation, 2026-10-15; fallback not reported
Metric: published percent
Original source · Retrieved 2026-10-15
Exact model and effort in the evaluator chart payload.
Terminal-Bench 4.0 — AA harness: 57.0707070707%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Muse Wonder Full (max); Artificial Analysis independent evaluation, 2026-10-15; fallback not reported
Metric: published percent
Original source · Retrieved 2026-10-15
Exact model and effort in the evaluator chart payload.
SciCode: 62.6157407407%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Muse Wonder Full (max); Artificial Analysis independent evaluation, 2026-10-15; fallback not reported
Metric: published percent
Original source · Retrieved 2026-10-15
Exact model and effort in the evaluator chart payload.
HLE text-only — AA harness: 53.6144578313%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Muse Wonder Full (max); Artificial Analysis independent evaluation, 2026-10-15; fallback not reported
Metric: published percent
Original source · Retrieved 2026-10-15
Exact model and effort in the evaluator chart payload.
GDP.pdf — AA harness: 32.6%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Muse Wonder Full (max); Artificial Analysis independent evaluation, 2026-10-15; fallback not reported
Metric: published percent
Original source · Retrieved 2026-10-15
Exact model and effort in the evaluator chart payload.
CritPt: 32.8571428571%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Muse Wonder Full (max); Artificial Analysis independent evaluation, 2026-10-15; fallback not reported
Metric: published percent
Original source · Retrieved 2026-10-15
Exact model and effort in the evaluator chart payload.
AA-Omniscience: 30.15 native-points
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Muse Wonder Full (max); Artificial Analysis independent evaluation, 2026-10-15; fallback not reported
Metric: published native-points
Original source · Retrieved 2026-10-15
Restored on 2026-10-28: 30.6 recorded 2026-10-28 came from the evaluator's renamed AA-Omniscience (AA-Omniscience v1.1: Knowledge and Hallucination Benchmark), a different version, so this row keeps its earlier value. Exact model and effort in the evaluator chart payload.
AA-LCR v1.1: 83.6666666667%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Muse Wonder Full (max); Artificial Analysis independent evaluation, 2026-10-15; fallback not reported
Metric: published percent
Original source · Retrieved 2026-10-15
Exact model and effort in the evaluator chart payload.
AA Intelligence Index v4.3.2 (score): 55.8021199856 native-points
Not used in the score. Source category: independent. Review status: reviewed-independent.
Configuration: Muse Wonder Full (max); Artificial Analysis independent evaluation, 2026-10-15; fallback not reported
Metric: published native-points
Original source · Retrieved 2026-10-15
Restored on 2026-10-28: 49.727212 recorded 2026-10-28 came from the evaluator's renamed AA Intelligence Index v4.3.2 (score) (Artificial Analysis Intelligence Index v5.0), a different version, so this row keeps its earlier value. Reference composite only, never scored alongside its component evaluations.
DeepSWE v1.1: 79%
Contributes to the score. Developer-reported. Review status: reviewed-developer.
Configuration: Muse Wonder Full (max); Muse Wonder Full developer-reported result (meta); fallback not reported
Metric: published percent
Original source · Retrieved 2026-10-15
Developer-reported result entered in the newsroom. Same protocol as the catalogue row: the lab table reproduces cells we hold (opus-5-5 74.2, fable-5-1 67.4). Value read from an image of the lab table. Value read from the scorecard image embedded on page 3 of the PDF model card. Per section 3, the Muse Wonder Full DeepSWE v1.1 result is Datacurve third-party (pass@1).
Finance Agent v2 — Vals: 59.1%
Contributes to the score. Developer-reported. Review status: reviewed-developer.
Configuration: Muse Wonder Full (max); Muse Wonder Full developer-reported result (meta); fallback not reported
Metric: published percent
Original source · Retrieved 2026-10-15
Developer-reported result entered in the newsroom. Same protocol as the catalogue row: the lab table reproduces cells we hold (muse-wonder 56.6, fable-5-1 58.9). Value read from an image of the lab table. Value read from the scorecard image embedded on page 3 of the PDF model card. Per section 3, the result source is Vals AI and the metric is private-test-set accuracy.
BioMysteryBench v1 — Vals: 73.33%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Muse Wonder Full (max); Vals AI BioMysteryBench harness; provider Meta; fallback not reported
Metric: published percent
Original source · Retrieved 2026-10-15
Vals AI model page, headline accuracy 73.33 ± 0.64, rank 8/27. Vals notes some benchmarks may use a different provider or parameters.
Code Migration †: 61.03%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Muse Wonder Full (max); Vals AI Code Migration harness; provider Meta; fallback not reported
Metric: published percent
Original source · Retrieved 2026-10-15
Vals AI model page, headline accuracy 61.03 ± 4.42, rank 7/77. Vals notes some benchmarks may use a different provider or parameters.
Excel Modeling Benchmark — Vals: 71.92%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Muse Wonder Full (max); Vals AI EMB harness; provider Meta; fallback not reported
Metric: published percent
Original source · Retrieved 2026-10-15
Vals AI model page, headline accuracy 71.92 ± 2.66, rank 9/74. Vals notes some benchmarks may use a different provider or parameters.
Harvey LAB (Vals): 21.67%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Muse Wonder Full (max); Vals AI Harvey's Legal Agent Benchmark harness; provider Meta; fallback not reported
Metric: published percent
Original source · Retrieved 2026-10-15
Vals AI model page, headline accuracy 21.67 ± 3.54, rank 4/79. Vals notes some benchmarks may use a different provider or parameters.
LegalBench — Vals: 87.42%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Muse Wonder Full (max); Vals AI LegalBench harness; provider Meta; fallback not reported
Metric: published percent
Original source · Retrieved 2026-10-15
Vals AI model page, headline accuracy 87.42 ± 0.41, rank 4/151. Vals notes some benchmarks may use a different provider or parameters.
Legal Research Bench — Vals: 53.37%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Muse Wonder Full (max); Vals AI Legal Research Bench harness; provider Meta; fallback not reported
Metric: published percent
Original source · Retrieved 2026-10-15
Vals AI model page, headline accuracy 53.37 ± 3.47, rank 5/78. Vals notes some benchmarks may use a different provider or parameters.
MedCode — Vals: 49.93%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Muse Wonder Full (max); Vals AI MedCode harness; provider Meta; fallback not reported
Metric: published percent
Original source · Retrieved 2026-10-15
Vals AI model page, headline accuracy 49.93 ± 2.18, rank 18/107. Vals notes some benchmarks may use a different provider or parameters.
MedScribe — Vals: 89.33%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Muse Wonder Full (max); Vals AI MedScribe harness; provider Meta; fallback not reported
Metric: published percent
Original source · Retrieved 2026-10-15
Vals AI model page, headline accuracy 89.33 ± 2.0, rank 8/109. Vals notes some benchmarks may use a different provider or parameters.
ProofBench v1.1 — Vals: 94%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Muse Wonder Full (max); Vals AI ProofBench v1.1 harness; provider Meta; fallback not reported
Metric: published percent
Original source · Retrieved 2026-10-15
Vals AI model page, headline accuracy 94.0 ± 2.39, rank 9/50. Vals notes some benchmarks may use a different provider or parameters.
Public Benefits Bench — Vals: 73.75%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Muse Wonder Full (max); Vals AI Public Benefits Bench v1.1 harness; provider Meta; fallback not reported
Metric: published percent
Original source · Retrieved 2026-10-15
Vals AI model page, headline accuracy 73.75 ± 1.15, rank 3/50. Vals notes some benchmarks may use a different provider or parameters.
Vibe Code Bench v1.1 — Vals: 86%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Muse Wonder Full (max); Vals AI Vibe Code Bench v1.1 harness; provider Meta; fallback not reported
Metric: published percent
Original source · Retrieved 2026-10-15
Vals AI model page, headline accuracy 86.0 ± 1.93, rank 12/112. Vals notes some benchmarks may use a different provider or parameters.
Cite this model record
The Latent. “Muse Wonder Full: score and benchmark evidence.” 2026-10-28. Release tli-2026-v1.0-2026-10-28-refresh-45.
Download release data and definitions (JSON) · Data usage terms