Score · relative 0–100: 52.5
Rank: 30
Status: Complete 3/3
Rank change: 0
Players generated: 2026-07-24 – 2026-10-14
Match evidence updated: 2026-10-26
Results published: 2026-10-26

Compare in charts View player programs Back to leaderboard

Scores by reasoning setting

High (Intensive group, tested and included)64.6
Medium (Balanced group, tested and included)52.5
Minimal (Baseline group, tested and included)40.4

Every tested setting is included in the overall score.

Available settings and grouping

Available settings

MinimalLowMediumHigh

Checked on 2026-07-24 · Provider documentation

How settings are grouped

Baseline: Minimal — tested and included
Balanced: Medium — tested and included
Intensive: High — tested and included

Scores by game

Go53.2
Chess1 of 6 failed38.0
Shogi58.3
Xiangqi48.9
Hex67.1
International Draughts63.4
Game of the Amazons49.1
Tumbleweed1 of 6 failed41.7

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 (64.6).

Across included settings, the highest game score is Hex (67.1); the lowest is Chess (38.0). 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: google/gemini-3.6-flash
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: 65536 (5 of 131 indexed attempts)

Generation evidence

Minimal: 16 contributing player programs across 8 games
Medium: 14 contributing player programs across 8 games
High: 16 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.

Game	Reasoning group	Programs	Working-program strength	Raw rating	Combined record (W / D / L)
Go	Intensive	2 working of 2 attempts	55.8 program scores 70.5 · 41.1	1530.8 program range 1384–1677	431 / 248 / 353
Go	Balanced	2 working of 2 attempts	72.4 program scores 93.5 · 51.2	1784.4 program range 1481–2088	597 / 269 / 166
Go	Baseline	2 working of 2 attempts	31.5 program scores 33.4 · 29.6	1287.4 program range 1267–1307	215 / 184 / 633
Chess	Intensive	2 working of 2 attempts	55.0 program scores 63.6 · 46.4	1512.8 program range 1444–1582	224 / 575 / 161
Chess	Balanced	1 working of 2 attempts	37.5	1374.5	5 / 335 / 140
Chess	Baseline	2 working of 2 attempts	40.3 program scores 40.6 · 40.1	1396.8 program range 1395–1399	6 / 731 / 223
Shogi	Intensive	2 working of 2 attempts	69.1 program scores 69.7 · 68.4	1614.8 program range 1607–1622	621 / 52 / 316
Shogi	Balanced	2 working of 2 attempts	49.3 program scores 66.1 · 32.6	1427.9 program range 1274–1581	413 / 79 / 496
Shogi	Baseline	2 working of 2 attempts	56.5 program scores 61.5 · 51.5	1487.6 program range 1441–1534	432 / 175 / 381
Xiangqi	Intensive	2 working of 2 attempts	63.1 program scores 72.0 · 54.2	1585.2 program range 1508–1663	377 / 469 / 150
Xiangqi	Balanced	2 working of 2 attempts	41.3 program scores 44.1 · 38.5	1407.8 program range 1386–1430	88 / 615 / 293
Xiangqi	Baseline	2 working of 2 attempts	42.4 program scores 44.9 · 39.9	1415.8 program range 1396–1435	61 / 689 / 246
Hex	Intensive	2 working of 2 attempts	83.9 program scores 91.1 · 76.6	2066.9 program range 1838–2296	879 / 0 / 181
Hex	Balanced	2 working of 2 attempts	70.0 program scores 89.2 · 50.7	1831.0 program range 1446–2216	730 / 0 / 330
Hex	Baseline	2 working of 2 attempts	47.6 program scores 50.1 · 45.1	1407.6 program range 1377–1439	488 / 0 / 572
International Draughts	Intensive	2 working of 2 attempts	82.6 program scores 91.3 · 73.9	2019.4 program range 1826–2213	804 / 14 / 186
International Draughts	Balanced	2 working of 2 attempts	83.8 program scores 85.5 · 82.1	2023.7 program range 1987–2060	800 / 47 / 157
International Draughts	Baseline	2 working of 2 attempts	23.9 program scores 31.4 · 16.3	1018.0 program range 898–1138	168 / 76 / 760
Game of the Amazons	Intensive	2 working of 2 attempts	51.4 program scores 52.2 · 50.6	1464.8 program range 1454–1475	519 / 0 / 541
Game of the Amazons	Balanced	2 working of 2 attempts	48.5 program scores 58.0 · 39.0	1430.1 program range 1311–1550	489 / 0 / 571
Game of the Amazons	Baseline	2 working of 2 attempts	47.5 program scores 59.8 · 35.2	1417.7 program range 1262–1573	479 / 0 / 581
Tumbleweed	Intensive	2 working of 2 attempts	55.9 program scores 74.6 · 37.2	1511.9 program range 1248–1776	504 / 134 / 438
Tumbleweed	Balanced	1 working of 2 attempts	71.3	1710.0	343 / 47 / 148
Tumbleweed	Baseline	2 working of 2 attempts	33.7 program scores 35.7 · 31.7	1210.6 program range 1189–1232	130 / 410 / 536

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

2 model-code generation failures; 5 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

High: 75% / 25% / 0% · n 16Medium: 88% / 0% / 12% · n 16Minimal: 75% / 25% / 0% · n 16 All three benchmark settings tested

Generation cost

$0.25average estimated cost per game and reasoning setting

Median: $0.25
Range: $0.04–$0.53
Known cost subtotal (incomplete or unverified charges): $13.08
Cost data: 24 combinations · 52 runs
Price source: Provider-reported
Price list: Not recorded for this older data
Average output: 32k tokens
Output range: 2.1k–100k tokens
Runs with output-token data: 52

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

Gemini 3.6 Flash