- Score · relative 0–100
- 32.7
- Rank
- 40
- Status
- Complete 3/3
- Players generated
- 2026-09-30
- Match evidence updated
- 2026-10-06
- Results published
Scores by reasoning setting
Available settings and grouping
Available settings
Checked on 2026-09-30 · Provider documentation
How settings are grouped
- Baseline: Low — tested and included
- Balanced: Medium — tested and included
- Intensive: High — tested and included
Scores by game
What this result shows
Strongest included setting: High (49.3).
Across included settings, the highest game score is Game of the Amazons (48.8); the lowest is Chess (9.2). 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-3.1-pro-preview
- Provider route (catalog)
- 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 (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; the per-game counts below keep unique code identities separate from admitted generation runs.
| Game | Setting | Unique programs | Admitted generation runs | Program score range |
|---|---|---|---|---|
| Chess | high | 1 | 1 | 9.0–9.0 |
| Chess | low | 1 | 1 | 8.5–8.5 |
| Chess | medium | 1 | 1 | 10.2–10.2 |
| Game of the Amazons | high | 1 | 1 | 60.6–60.6 |
| Game of the Amazons | low | 1 | 1 | 28.6–28.6 |
| Game of the Amazons | medium | 1 | 1 | 57.2–57.2 |
| Go | high | 1 | 1 | 57.3–57.3 |
| Go | low | 1 | 1 | 36.0–36.0 |
| Go | medium | 1 | 1 | 33.0–33.0 |
| Hex | high | 1 | 1 | 74.1–74.1 |
| Hex | low | 1 | 1 | 20.5–20.5 |
| Hex | medium | 1 | 1 | 23.7–23.7 |
| International Draughts | high | 1 | 1 | 29.2–29.2 |
| International Draughts | low | 1 | 1 | 25.3–25.3 |
| International Draughts | medium | 1 | 1 | 34.5–34.5 |
| Shogi | high | 1 | 1 | 40.9–40.9 |
| Shogi | low | 1 | 1 | 12.8–12.8 |
| Shogi | medium | 1 | 1 | 39.8–39.8 |
| Tumbleweed | high | 1 | 1 | 62.3–62.3 |
| Tumbleweed | low | 1 | 1 | 17.5–17.5 |
| Tumbleweed | medium | 1 | 1 | 17.7–17.7 |
| Xiangqi | high | 1 | 1 | 61.1–61.1 |
| Xiangqi | low | 1 | 1 | 9.7–9.7 |
| Xiangqi | medium | 1 | 1 | 14.4–14.4 |
Program ranges describe the retained competitors on this release scale, not a statistical estimate of a fresh generation.
Failures and budgets
0 model-code generation failures; 0 provider/infrastructure failures excluded from the generation denominator. Runtime player faults are separate: 0 recorded unilateral faults across contributing program rows, excluded from 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: 100% / 0% / 0% · n 8Low: 88% / 12% / 0% · n 8 All three benchmark settings testedGeneration cost
$0.15average estimated cost per game and reasoning setting
- Median
- $0.08
- Range
- $0.04–$0.47
- Estimated total for recorded runs
- $3.60
- Cost data
- 24 combinations · 24 runs
- Price source
- Standard list-price estimate
- Price list
- boardgame-list-prices-2026-09-08
- Average output
- 11k tokens
- Output range
- 1.6k–36k 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