DuelLabGameBench 2
Menu

Model results · Google

Gemini 4 Argon

Rank 5 of 66 in Overall Models. All three benchmark settings tested.

Score · relative 0–100
83.3
Rank
5
Status
Complete 3/3
Rank change
0
Players generated
2026-10-21
Match evidence updated
2026-10-24
Results published
Compare in charts View player programs Back to leaderboard

Scores by reasoning setting

  • High (Intensive group, tested and included)90.1
  • Medium (Balanced group, tested and included)78.7
  • Low (Baseline group, tested and included)81.0

Every tested setting is included in the overall score.

Available settings and grouping

Available settings

LowMediumHigh

Checked on 2026-10-21 · Provider documentation

How settings are grouped

  • Baseline: Low — tested and included
  • Balanced: Medium — tested and included
  • Intensive: High — tested and included

Scores by game

  • Go93.0
  • Chess79.1
  • Shogi86.0
  • Xiangqi72.1
  • Hex91.7
  • International Draughts93.6
  • Game of the Amazons73.9
  • Tumbleweed76.9

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: High (90.1).

Across included settings, the highest game score is International Draughts (93.6); the lowest is Xiangqi (72.1). 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
gemini-4-argon
Provider route (catalog)
google
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 (28 of 28 indexed attempts)

Generation evidence

  • Low: 8 contributing player programs across 8 games
  • Medium: 8 contributing player programs across 8 games
  • High: 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 attempt95.32163.4489 / 0 / 27
GoBalanced1 working of 1 attempt91.12012.9464 / 0 / 52
GoBaseline1 working of 1 attempt92.52053.6472 / 0 / 44
ChessIntensive1 working of 1 attempt94.92087.1433 / 45 / 2
ChessBalanced1 working of 1 attempt65.31597.0226 / 159 / 95
ChessBaseline1 working of 1 attempt77.11717.7292 / 144 / 44
ShogiIntensive1 working of 1 attempt92.32080.3430 / 44 / 20
ShogiBalanced1 working of 1 attempt80.01766.4332 / 100 / 62
ShogiBaseline1 working of 1 attempt85.81880.2404 / 21 / 70
XiangqiIntensive1 working of 1 attempt71.41656.9312 / 71 / 115
XiangqiBalanced1 working of 1 attempt68.91633.0245 / 180 / 73
XiangqiBaseline1 working of 1 attempt76.11707.4298 / 148 / 52
HexIntensive1 working of 1 attempt96.32606.2512 / 0 / 18
HexBalanced1 working of 1 attempt90.42266.0477 / 0 / 53
HexBaseline1 working of 1 attempt88.42182.3465 / 0 / 65
International DraughtsIntensive1 working of 1 attempt97.62526.6486 / 8 / 8
International DraughtsBalanced1 working of 1 attempt92.72259.9449 / 25 / 28
International DraughtsBaseline1 working of 1 attempt90.42186.2441 / 15 / 46
Game of the AmazonsIntensive1 working of 1 attempt84.31991.8446 / 0 / 84
Game of the AmazonsBalanced1 working of 1 attempt77.91851.0409 / 0 / 121
Game of the AmazonsBaseline1 working of 1 attempt59.41568.5304 / 0 / 226
TumbleweedIntensive1 working of 1 attempt88.72119.3471 / 0 / 67
TumbleweedBalanced1 working of 1 attempt63.61575.6323 / 0 / 215
TumbleweedBaseline1 working of 1 attempt78.51858.9409 / 0 / 129

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: 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: 63% / 37% / 0% · n 8Medium: 88% / 12% / 0% · n 8Low: 100% / 0% / 0% · n 8 All three benchmark settings tested

Generation cost

$0.21average estimated cost per game and reasoning setting

Median
$0.09
Range
$0.06–$0.58
Estimated total for recorded runs
$4.97
Cost data
24 combinations · 24 runs
Price source
Standard list-price estimate
Price list
boardgame-list-prices-2026-10-09
Average output
18k tokens
Output range
4.6k–55k 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