DuelLabGameBench 2
Menu

Model results · Meta

Muse Wonder

Rank 11 of 66 in Overall Models. All three benchmark settings tested.

Score · relative 0–100
66.4
Rank
11
Status
Complete 3/3
Rank change
0
Players generated
2026-10-12
Match evidence updated
2026-10-24
Results published
Compare in charts View player programs Back to leaderboard

Scores by reasoning setting

  • XHigh (Intensive group, tested and included)78.8
  • Medium (Balanced group, tested and included)63.5
  • Minimal (Baseline group, tested and included)56.8

Every tested setting is included in the overall score.

Available settings and grouping

Available settings

MinimalLowMediumHighXHighMax

Checked on 2026-10-12 · Provider documentation

How settings are grouped

  • Baseline: Minimal — tested and included
  • Balanced: Medium — tested and included
  • Intensive: XHigh — tested and included

Scores by game

  • Go69.1
  • Chess55.3
  • Shogi45.6
  • Xiangqi58.4
  • Hex87.9
  • International Draughts76.9
  • Game of the Amazons71.9
  • Tumbleweed65.8

Each game counts every generation attempt, a failed one as 0 (“1 of 2 failed” marks games with failed attempts). Scores are relative within each game, across the included settings; they do not compare game difficulty.

What this result shows

Strongest included setting: XHigh (78.8).

Across included settings, the highest game score is Hex (87.9); the lowest is Shogi (45.6). Each game counts every generation attempt, a failed one as 0. These are relative scores within each game, not direct comparisons of difficulty.

Match evidence measures these retained programs. It does not establish how a fresh generation will perform. Independent-generation variation is not estimated in this release.

Test protocol and evidence limits
Model identifier
meta/muse-wonder
Provider route (catalog)
openrouter
Prompt identity / profile
Not retained in this result set / Not retained in this result set
Prompt version
Not retained in this result set
Generation policy
Generation-quality admission policy (see methodology)
Requested output-token limits
943718 (30 of 30 indexed attempts)

Generation evidence

  • Minimal: 8 contributing player programs across 8 games
  • Medium: 8 contributing player programs across 8 games
  • XHigh: 8 contributing player programs across 8 games

Counts describe contributing programs, not independent replications: repairs and resends are not new independent samples. Independent-generation lineage is not retained.

GameReasoning groupProgramsWorking-program strengthRaw ratingCombined record (W / D / L)
GoIntensive1 working of 1 attempt85.11881.0428 / 5 / 83
GoBalanced1 working of 1 attempt53.31500.4262 / 4 / 250
GoBaseline1 working of 1 attempt69.01660.1263 / 164 / 89
ChessIntensive1 working of 1 attempt66.61609.1177 / 270 / 33
ChessBalanced1 working of 1 attempt56.21520.6115 / 293 / 72
ChessBaseline1 working of 1 attempt43.01417.812 / 373 / 95
ShogiIntensive1 working of 1 attempt83.01820.9356 / 85 / 53
ShogiBalanced1 working of 1 attempt36.91313.0113 / 103 / 278
ShogiBaseline1 working of 1 attempt17.11107.229 / 89 / 376
XiangqiIntensive1 working of 1 attempt81.91779.4370 / 64 / 64
XiangqiBalanced1 working of 1 attempt39.61393.982 / 214 / 202
XiangqiBaseline1 working of 1 attempt53.91505.2106 / 307 / 85
HexIntensive1 working of 1 attempt93.12395.1493 / 0 / 37
HexBalanced1 working of 1 attempt85.32071.1447 / 0 / 83
HexBaseline1 working of 1 attempt85.22065.3446 / 0 / 84
International DraughtsIntensive1 working of 1 attempt84.52039.8400 / 32 / 70
International DraughtsBalanced1 working of 1 attempt77.11886.7360 / 30 / 112
International DraughtsBaseline1 working of 1 attempt69.01735.2311 / 38 / 152
Game of the AmazonsIntensive1 working of 1 attempt74.11783.0387 / 0 / 143
Game of the AmazonsBalanced1 working of 1 attempt87.52088.2465 / 0 / 65
Game of the AmazonsBaseline1 working of 1 attempt54.01498.3274 / 0 / 256
TumbleweedIntensive1 working of 1 attempt62.01551.5275 / 79 / 184
TumbleweedBalanced1 working of 1 attempt72.01723.2358 / 25 / 155
TumbleweedBaseline1 working of 1 attempt63.31571.3284 / 75 / 179

Each row covers all the model's programs in one game and setting. Strength and raw rating are attempt-weighted means; under them, each program's score and the lowest and highest rating describe the programs (not a confidence interval). The combined record adds up their matches: evidence volume, not strength.

Failures and budgets

0 model-code generation failures; 0 provider/infrastructure failures excluded from the generation denominator. Runtime player faults are separate: 1 recorded one-sided faults across contributing program rows. A crash or illegal move pinned on one program also counts as that program's loss in its match record; other faults, such as timeouts, are left out of rating evidence.

Repair/resend outcomes and available costs are included. Exact retry ceilings and execution budgets are not retained in this result set; see the methodology for the public policy. Do not infer a measured setting from a current provider catalog.

Player program results

XHigh: 100% / 0% / 0% · n 8Medium: 75% / 25% / 0% · n 8Minimal: 88% / 12% / 0% · n 8 All three benchmark settings tested

Generation cost

$0.29average estimated cost per game and reasoning setting

Median
$0.21
Range
$0.10–$0.68
Provider-reported total
$6.87
Cost data
24 combinations · 24 runs
Price source
Provider-reported
Price list
boardgame-list-prices-2026-10-09
Average output
12k tokens
Output range
3.3k–30k tokens
Runs with output-token data
24
How cost is measured

Comparison cost gives equal weight to each model, game, and reasoning-setting combination. Recorded generation attempts include repairs and resends when cost evidence is available; runtime compute and service operation are excluded. Cost and output-token counts use separate telemetry denominators. A list-price estimate is not actual expenditure; mixed evidence is not a uniform standard-price comparison.

Test protocol

  • Combines this model's tested reasoning settings
  • Player-program generation; retained protocol details are under “Test protocol and evidence limits” above