Scores by reasoning setting

Enabled (Balanced group, tested and included)30.6
Enablednot included (Balanced group, tested, not included)30.6
None (Baseline group, tested and included)18.8
Nonenot included (Baseline group, tested, not included)19.0
Nonenot included (Baseline group, tested, not included)30.2

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

Available settings and grouping

Available settings

NoneEnabled

Checked on 2026-10-09 · Provider documentation

How settings are grouped

Baseline: None — tested and included
Balanced: Enabled — tested and included
Intensive: Not available for this model

Scores by game

Go27.6
Chess1 of 3 failed28.9
Shogi29.0
Xiangqi1 of 3 failed31.1
Hex2 of 2 failed0.0
International Draughts14.6
Game of the Amazons2 of 4 failed18.0
Tumbleweed33.4

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: Enabled (30.6).

Across included settings, the highest game score is Tumbleweed (33.4); the lowest is Hex (0.0). 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: inclusionai/ling-3.0-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: 32768 (4 of 98 indexed attempts)

Generation evidence

None: 11 contributing player programs across 6 games
Enabled: 6 contributing player programs across 5 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	Balanced	1 working of 1 attempt	19.5	1145.0	29 / 129 / 358
Go	Baseline	2 working of 2 attempts	35.7 program scores 40.9 · 30.4	1329.2 program range 1276–1382	252 / 194 / 586
Chess	Balanced	1 working of 1 attempt	37.5	1374.5	4 / 337 / 139
Chess	Baseline	1 working of 2 attempts	40.6	1398.8	2 / 370 / 108
Shogi	Baseline	2 working of 2 attempts	29.0 program scores 32.8 · 25.1	1238.9 program range 1201–1276	101 / 309 / 578
Xiangqi	Balanced	1 working of 1 attempt	40.0	1397.1	21 / 340 / 137
Xiangqi	Baseline	1 working of 2 attempts	44.7	1433.7	35 / 358 / 105
International Draughts	Balanced	1 working of 1 attempt	19.9	958.5	83 / 3 / 416
International Draughts	Baseline	2 working of 2 attempts	9.4 program scores 17.6 · 1.2	627.6 program range 335–920	75 / 5 / 924
Game of the Amazons	Balanced	2 working of 2 attempts	36.0 program scores 38.3 · 33.7	1271.6 program range 1242–1301	354 / 0 / 706
Tumbleweed	Baseline	3 working of 3 attempts	33.4 program scores 36.8 · 33.1 · 30.2	1207.0 program range 1173–1243	249 / 497 / 868

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

6 model-code generation failures; 10 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

Enabled: 17% / 83% / 0% · n 6None: 41% / 24% / 35% · n 17 No Overall rankIntensive not supportedMany programs needed repairMany programs failed

Generation cost

$0.000204average estimated cost per game and reasoning setting

Median: $0.00
Range: $0.00–$0.0022
Known cost subtotal (incomplete or unverified charges): $0.0098
Cost data: 16 combinations · 34 runs
Price source: Provider-reported
Price list: boardgame-list-prices-2026-10-09
Average output: 33k tokens
Output range: 2.0k–92k tokens
Runs with output-token data: 34

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

Ling 3.0 Flash