DuelLabGameBench 2
Menu

Model results · Xiaomi

MiMo-V2.5

No Overall rank: the model offers 2 of 3 reasoning settings.

Score · relative 0–100
30.7
Rank
—
Status
Provisional
Players generated
2026-07-01 – 2026-07-03
Match evidence updated
2026-10-26
Results published
Compare in charts View player programs Back to leaderboard

Scores by reasoning setting

  • Highnot included (Intensive group, tested, not included)35.9
  • Enabled (Balanced group, tested and included)30.2
  • Mediumnot included (Balanced group, tested, not included)30.2
  • Reasoningnot included (Balanced group, tested, not included)34.1
  • Mediumnot included (Balanced group, tested, not included)29.4
  • None (Baseline group, tested and included)26.1
  • Reasoningnot included (Baseline group, tested, not included)24.6
  • Reasoningnot included (Baseline group, tested, not included)27.0

Settings marked “not included” were tested but are left out of the overall score.

Available settings and grouping

Available settings

Reasoning disabledReasoning enabled

Checked on 2026-07-21 · Provider documentation

How settings are grouped

  • Baseline: Reasoning disabled — tested and included
  • Balanced: Reasoning enabled — tested and included
  • Intensive: Not available for this model

Scores by game

  • Go32.0
  • Chess2 of 7 failed33.4
  • Shogi2 of 3 failed15.2
  • Xiangqi40.1
  • Hex26.1
  • International Draughts34.6
  • Game of the Amazons1 of 5 failed22.2
  • Tumbleweed2 of 7 failed22.2

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: Reasoning enabled (30.2).

Across included settings, the highest game score is Xiangqi (40.1); the lowest is Shogi (15.2). 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
xiaomi/mimo-v2.5
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
Not retained (0 of 32 indexed attempts)

Generation evidence

  • Reasoning disabled: 11 contributing player programs across 7 games
  • Reasoning enabled: 7 contributing player programs across 7 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 attempt23.51197.26 / 215 / 295
GoBalanced2 working of 2 attempts29.9 program scores 32.6 · 27.21270.0 program range 1241–1299225 / 132 / 675
GoBaseline3 working of 3 attempts42.7 program scores 57.8 · 35.6 · 34.81398.4 program range 1322–1544485 / 292 / 771
ChessIntensive1 working of 1 attempt44.11426.011 / 385 / 84
ChessBalanced3 working of 3 attempts42.4 program scores 43.8 · 42.0 · 41.51413.1 program range 1406–142432 / 1110 / 298
ChessBaseline1 working of 3 attempts41.21403.86 / 368 / 106
ShogiBaseline1 working of 3 attempts45.71389.9116 / 181 / 197
XiangqiIntensive1 working of 1 attempt39.61393.936 / 306 / 156
XiangqiBalanced3 working of 3 attempts38.8 program scores 40.4 · 38.2 · 37.81387.9 program range 1380–140090 / 932 / 472
XiangqiBaseline2 working of 2 attempts42.0 program scores 42.1 · 41.91413.1 program range 1412–141497 / 610 / 289
HexBalanced2 working of 2 attempts24.4 program scores 27.8 · 21.01084.2 program range 1029–1140247 / 0 / 813
HexBaseline3 working of 3 attempts27.7 program scores 50.7 · 17.8 · 14.71106.6 program range 906–1446422 / 0 / 1168
International DraughtsIntensive1 working of 1 attempt49.11405.0214 / 20 / 268
International DraughtsBalanced1 working of 1 attempt21.9991.487 / 13 / 402
International DraughtsBaseline3 working of 3 attempts33.0 program scores 61.1 · 19.4 · 18.41160.7 program range 934–1597442 / 12 / 1052
Game of the AmazonsBalanced2 working of 2 attempts30.7 program scores 37.9 · 23.61198.3 program range 1101–1296300 / 0 / 760
Game of the AmazonsBaseline2 working of 3 attempts20.4 program scores 29.3 · 11.61028.5 program range 874–1183195 / 0 / 865
TumbleweedIntensive1 working of 1 attempt23.11091.723 / 182 / 333
TumbleweedBalanced2 working of 3 attempts34.7 program scores 50.6 · 18.81217.7 program range 1037–1398212 / 269 / 595
TumbleweedBaseline2 working of 3 attempts30.6 program scores 33.9 · 27.41177.1 program range 1141–121377 / 454 / 545

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

7 model-code generation failures; 12 provider/infrastructure failures excluded from the generation denominator. Runtime player faults are separate: 0 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

High: 0% / 100% / 0% · n 5Reasoning enabled: 6% / 88% / 6% · n 16Reasoning disabled: 52% / 22% / 26% · n 23 No Overall rankMany programs needed repairMany programs failed

Generation cost

$0.02average estimated cost per game and reasoning setting

Median
$0.01
Range
$0.0017–$0.06
Known cost subtotal (incomplete or unverified charges)
$0.75
Cost data
24 combinations · 56 runs
Price source
Provider-reported
Price list
Not recorded for this older data
Average output
43k tokens
Output range
1.6k–131k tokens
Runs with output-token data
17
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