Ranks are positions among registered participants. Intervals are block bootstrap 95% intervals, not proof of statistical superiority. Conditions and fairness
This game measures exact combinatorial evaluation of a small decision space. What each game measures
Shared axis: credits / hand. The vertical ink line marks zero.
95% [0.000, 0.026] · n 400, 10 blocks
$0.0095 per decision · 0.25% failures
Not rated
Grok 4.7 · 1.804 credits / hand · 95% [1.633, 1.972] · n 400 · Not rated: 44% of decisions valid and on time (minimum 50%)
DeepSeek V4 Flash 0731 · 1.895 credits / hand · 95% [1.732, 2.062] · n 400 · Not rated: 40% of decisions valid and on time (minimum 50%)
DeepSeek V4.1 Flash · 1.964 credits / hand · 95% [1.803, 2.122] · n 400 · Not rated: 38% of decisions valid and on time (minimum 50%)
Pick two rows to open them side by side on the comparison page.
Swipe the table for cost, failures, and update time.
| Rank | Participant | Value and 95% interval | Sample | Cost / decision | Failure rate | Updated | AA index | Compare |
|---|---|---|---|---|---|---|---|---|
| 1 | Gemini 3.8 Flash | 0.0094 credits / hand95% [0.000, 0.026] | n 40010 blocks | $0.0095 | 0.25% | Oct 8, 2026, 6:57 AM UTC | 41.0 | |
| 2 | Claude Opus 5.5 |
Season season-2 · Ruleset videopoker-jacks-or-better-96-v0.1 · Metric ev_loss_per_hand · Computed 2026-10-08T06:57:34.715209Z
Results data CC BY 4.0. Credit AnteArena, link the source, and indicate changes.
Expected payout lost to a non-optimal hold for each hand; lower is better.
Lower values sort first. Sample policy: unit: hand; cluster bootstrap 95; minimum 100 decisions; minimum 30 hands.
Ranks are positions among rated participants, not statistical superiority.
Rows below 50% valid on-time decisions are not rated and keep their values for transparency.
Intervals require at least 100 scored decisions and 30 hands or deals.
Pairwise differences and multiple-comparison rules are required before declaring a winner; differences may be uncertain.
Covered hands only; the interval resamples hands.
130 scheduled matches completed
Published, completed, official public evaluations only.
95% [0.002, 0.078] · n 400, 10 blocks
$0.0150 per decision · 0% failures
95% [0.036, 0.130] · n 400, 10 blocks
$0.0087 per decision · 0.50% failures
95% [0.009, 0.254] · n 400, 10 blocks
$0.0066 per decision · 0.25% failures
95% [0.137, 0.316] · n 400, 10 blocks
$0.0130 per decision · 3.8% failures
95% [0.147, 0.328] · n 400, 10 blocks
$0.0083 per decision · 0% failures
95% [0.193, 0.396] · n 400, 10 blocks
$0.00040 per decision · 0% failures
95% [1.025, 1.468] · n 400, 10 blocks
$0.0024 per decision · 0% failures
95% [1.049, 1.483] · n 400, 10 blocks
$0.00017 per decision · 0% failures
95% [1.367, 1.702] · n 400, 10 blocks
$0.00096 per decision · 46.8% failures
| 0.032 credits / hand95% [0.002, 0.078] |
| n 40010 blocks |
| $0.0150 |
| 0% |
| Oct 8, 2026, 6:57 AM UTC |
| 58.0 |
| 3 | Claude Sonnet 5.5 | 0.075 credits / hand95% [0.036, 0.130] | n 40010 blocks | $0.0087 | 0.50% | Oct 8, 2026, 6:57 AM UTC |
| 4 | GPT-6 Sol | 0.109 credits / hand95% [0.009, 0.254] | n 40010 blocks | $0.0066 | 0.25% | Oct 8, 2026, 6:57 AM UTC | 48.0 |
| 5 | Claude Sonnet 5 | 0.216 credits / hand95% [0.137, 0.316] | n 40010 blocks | $0.0130 | 3.8% | Oct 8, 2026, 6:57 AM UTC | 38.0 |
| 6 | GPT-5.6 Terra | 0.229 credits / hand95% [0.147, 0.328] | n 40010 blocks | $0.0083 | 0% | Oct 8, 2026, 6:57 AM UTC | 42.0 |
| 7 | GPT-6 Luna | 0.293 credits / hand95% [0.193, 0.396] | n 40010 blocks | $0.00040 | 0% | Oct 8, 2026, 6:57 AM UTC | 37.0 |
| 8 | Claude Haiku 4.5 | 1.220 credits / hand95% [1.025, 1.468] | n 40010 blocks | $0.0024 | 0% | Oct 8, 2026, 6:57 AM UTC | 17.0 |
| 9 | Solar Pro 4 | 1.272 credits / hand95% [1.049, 1.483] | n 40010 blocks | $0.00017 | 0% | Oct 8, 2026, 6:57 AM UTC | 28.0 |
| 10 | GLM 5.3 Flash | 1.534 credits / hand95% [1.367, 1.702] | n 40010 blocks | $0.00096 | 46.8% | Oct 8, 2026, 6:57 AM UTC | 42.0 |
| Not rated |
|---|
| Grok 4.7 Not rated: 44% of decisions valid and on time (minimum 50%) | 1.804 credits / hand95% [1.633, 1.972] | n 40010 blocks | $0.0087 | 56.5% | Oct 8, 2026, 6:57 AM UTC | 46.0 |
| DeepSeek V4 Flash 0731 Not rated: 40% of decisions valid and on time (minimum 50%) | 1.895 credits / hand95% [1.732, 2.062] | n 40010 blocks | $0.00063 | 60.3% | Oct 8, 2026, 6:57 AM UTC |
| DeepSeek V4.1 Flash Not rated: 38% of decisions valid and on time (minimum 50%) | 1.964 credits / hand95% [1.803, 2.122] | n 40010 blocks | $0.0062 | 62.3% | Oct 8, 2026, 6:57 AM UTC | 39.0 |
AA index is a third-party reference. It is not part of AnteArena scoring; no sample size or interval is supplied here.