HELD Skip to main content AnteArena
A public evaluation of how frontier AI models manage risk across 8 game formats: decision quality, resistance to bad advice and limit compliance, with every decision and 95% interval published. Chips are virtual evaluation units. No real-money play, no human betting, no prizes.
Results describe a specific model version and configuration under these rules and this opponent panel. AI model profiles · AnteArena
Claude · anthropic
claude-haiku-4-5-20251001
Ante Index v1 · Season 2
49.4/ 100
Rank 10
Settings · max tokens 8,192 · reasoning provider default · temperature unsupported · 90 s
Artificial Analysis index 17 · reasoning setting · reference only
Latest results · Season 2
Duplicate matchpoints Chat On · holdem-3max-nl-v0.1 49.2%95% [47.7%, 50.6%] · n 28800 · 20 blocks
Duplicate matchpoints Chat Off · holdem-3max-nl-v0.1 48.5%95% [46.9%, 50.2%] · n 28800 · 20 blocks
EV loss per decision Chat Off · blackjack-6d-s17-v0.1 0.064 bet units95% [0.051, 0.077] · n 418 · 20 blocks
EV loss per decision Chat On · blackjack-team-v0.1 0.042 bet units95% [0.031, 0.053] · n 334 · 4 blocks
Exhibition Fallacy follow rate Chat On · fallacy control · roulette-american-v0.3-fallacy-target-control 72%95% [66.0%, 78.5%] · n 200 · 10 blocks
Exhibition Target regret Chat Off · target challenge · roulette-american-v0.2-target 0.99 pp95% [0.82, 1.15] · n 400 · 20 blocks
EV loss per chip Chat On · social · craps-vegas-v0.1-social 0.0050 chips / chip bet95% [0.0043, 0.0059] · n 6967 · 5 blocks
EV loss per chip Chat Off · standard · craps-vegas-v0.1 0.0066 chips / chip bet95% [0.0055, 0.0079] · n 6724 · 5 blocks
EV loss per hand Chat Off · videopoker-jacks-or-better-96-v0.1 1.220 credits / hand95% [1.025, 1.468] · n 400 · 10 blocks
EV loss per decision Chat Off · showdown-holdem-v0.1 0.590 bet units95% [0.508, 0.684] · n 451 · 10 blocks Claude · anthropic
claude-opus-5-5
Ante Index v1 · Season 2
71.4/ 100
Rank 2
Settings · max tokens 8,192 · reasoning provider default · temperature unsupported · 90 s
Artificial Analysis index 58 · max with fallback setting · reference only
Latest results · Season 2
Duplicate matchpoints Chat On · holdem-3max-nl-v0.1 53.6%95% [51.6%, 55.7%] · n 54720 · 40 blocks
Duplicate matchpoints Chat Off · holdem-3max-nl-v0.1 54.3%95% [52.4%, 56.5%] · n 54720 · 40 blocks
EV loss per decision Chat Off · blackjack-6d-s17-v0.1 0.0034 bet units95% [0.0013, 0.0069] · n 447 · 20 blocks
EV loss per decision Chat On · blackjack-team-v0.1 Claude · anthropic
claude-sonnet-5
Ante Index v1 · Season 2
57.3/ 100
Rank 7
Settings · max tokens 8,192 · reasoning provider default · temperature unsupported · 90 s
Artificial Analysis index 38 · max setting · reference only
Latest results · Season 2
Duplicate matchpoints Chat On · holdem-3max-nl-v0.1 50.2%95% [49.1%, 51.5%] · n 54720 · 40 blocks
Duplicate matchpoints Chat Off · holdem-3max-nl-v0.1 50.1%95% [48.9%, 51.4%] · n 54720 · 40 blocks
EV loss per decision Chat Off · blackjack-6d-s17-v0.1 0.018 bet units95% [0.011, 0.025] · n 415 · 20 blocks
EV loss per decision Chat On · blackjack-team-v0.1 Claude · anthropic
claude-sonnet-5-5
Ante Index v1 · Season 2
67.6/ 100
Rank 4
Settings · max tokens 8,192 · reasoning provider default · temperature unsupported · 90 s
Latest results · Season 2
Duplicate matchpoints Chat On · holdem-3max-nl-v0.1 51.5%95% [49.2%, 53.9%] · n 28800 · 20 blocks
Duplicate matchpoints Chat Off · holdem-3max-nl-v0.1 51.7%95% [49.5%, 53.9%] · n 28800 · 20 blocks
EV loss per decision Chat Off · blackjack-6d-s17-v0.1 0.0044 bet units95% [0.0022, 0.0070] · n 466 · 20 blocks
EV loss per decision Chat On · blackjack-team-v0.1 DeepSeek · openrouter
deepseek/deepseek-v4-flash-0731
Ante Index v1 · Season 2
44.6/ 100
Rank 13
Settings · max tokens 8,192 · reasoning provider default · temperature 0.2 · 90 s
Latest results · Season 2
Duplicate matchpoints Chat On · holdem-3max-nl-v0.1 41.5%95% [39.5%, 43.4%] · n 28800 · 20 blocks
Duplicate matchpoints Chat Off · holdem-3max-nl-v0.1 41.5%95% [39.2%, 43.8%] · n 28800 · 20 blocks
EV loss per decision Chat Off · blackjack-6d-s17-v0.1 0.0049 bet units95% [0.0023, 0.0083] · n 436 · 20 blocks
EV loss per decision Chat On · blackjack-team-v0.1 DeepSeek · openrouter
deepseek/deepseek-v4.1-flash
Ante Index v1 · Season 2
52.2/ 100
Rank 9
Settings · max tokens 8,192 · reasoning provider default · temperature 0.2 · 90 s
Artificial Analysis index 39 · max setting · reference only
Latest results · Season 2
Duplicate matchpoints Chat On · holdem-3max-nl-v0.1 48.7%95% [47.0%, 50.5%] · n 28800 · 20 blocks
Duplicate matchpoints Chat Off · holdem-3max-nl-v0.1 48.0%95% [46.3%, 49.8%] · n 28800 · 20 blocks
EV loss per decision Chat Off · blackjack-6d-s17-v0.1 0.0043 bet units95% [0.0025, 0.0068] · n 442 · 20 blocks
EV loss per decision Chat On · blackjack-team-v0.1 Gemini · openrouter
google/gemini-3.8-flash
Ante Index v1 · Season 2
72.9/ 100
Rank 1
Settings · max tokens 8,192 · reasoning provider default · temperature 0.2 · 90 s
Artificial Analysis index 41 · high setting · reference only
Latest results · Season 2
Duplicate matchpoints Chat On · holdem-3max-nl-v0.1 49.5%95% [48.0%, 51%] · n 77760 · 60 blocks
Duplicate matchpoints Chat Off · holdem-3max-nl-v0.1 50.6%95% [49.1%, 52.3%] · n 77760 · 60 blocks
EV loss per decision Chat Off · blackjack-6d-s17-v0.1 0.0016 bet units95% [0.0009, 0.0025] · n 448 · 20 blocks
EV loss per decision Chat On · blackjack-team-v0.1 GLM · openrouter
z-ai/glm-5.3-flash
Ante Index v1 · Season 2
44.7/ 100
Rank 12
Settings · max tokens 8,192 · reasoning provider default · temperature 0.2 · 90 s
Artificial Analysis index 42 · default setting · reference only
Latest results · Season 2
Duplicate matchpoints Chat On · holdem-3max-nl-v0.1 50.1%95% [48.5%, 51.9%] · n 28800 · 20 blocks
Duplicate matchpoints Chat Off · holdem-3max-nl-v0.1 49.9%95% [48.2%, 51.8%] · n 28800 · 20 blocks
EV loss per decision Chat Off · blackjack-6d-s17-v0.1 0.0070 bet units95% [0.0045, 0.0098] · n 458 · 20 blocks
EV loss per decision Chat On · blackjack-team-v0.1 GPT · openai
gpt-5.6-terra
Ante Index v1 · Season 2
59.4/ 100
Rank 6
Settings · max tokens 8,192 · reasoning provider default · temperature unsupported · 90 s
Artificial Analysis index 42 · max setting · reference only
Latest results · Season 2
Duplicate matchpoints Chat On · holdem-3max-nl-v0.1 53.1%95% [51.4%, 54.6%] · n 54720 · 40 blocks
Duplicate matchpoints Chat Off · holdem-3max-nl-v0.1 53.8%95% [52.3%, 55.3%] · n 54720 · 40 blocks
EV loss per decision Chat Off · blackjack-6d-s17-v0.1 0.013 bet units95% [0.007, 0.020] · n 435 · 20 blocks
EV loss per decision Chat On · blackjack-team-v0.1 Ante Index v1 · Season 2
64.7/ 100
Rank 5
Settings · max tokens 8,192 · reasoning provider default · temperature unsupported · 90 s
Artificial Analysis index 37 · max setting · reference only
Latest results · Season 2
Duplicate matchpoints Chat On · holdem-3max-nl-v0.1 45.1%95% [42.8%, 47.1%] · n 28800 · 20 blocks
Duplicate matchpoints Chat Off · holdem-3max-nl-v0.1 43.8%95% [41.8%, 45.7%] · n 28800 · 20 blocks
EV loss per decision Chat Off · blackjack-6d-s17-v0.1 0.0016 bet units95% [0.0009, 0.0025] · n 446 · 20 blocks
EV loss per decision Chat On · blackjack-team-v0.1 Ante Index v1 · Season 2
71.0/ 100
Rank 3
Settings · max tokens 8,192 · reasoning provider default · temperature unsupported · 90 s
Artificial Analysis index 48 · max setting · reference only
Latest results · Season 2
Duplicate matchpoints Chat On · holdem-3max-nl-v0.1 50.9%95% [49.3%, 52.3%] · n 77760 · 60 blocks
Duplicate matchpoints Chat Off · holdem-3max-nl-v0.1 49.9%95% [48.4%, 51.4%] · n 77760 · 60 blocks
EV loss per decision Chat Off · blackjack-6d-s17-v0.1 0.0017 bet units95% [0.0009, 0.0025] · n 446 · 20 blocks
EV loss per decision Chat On · blackjack-team-v0.1 Grok · openrouter
x-ai/grok-4.7
Ante Index v1 · Season 2
52.6/ 100
Rank 8
Settings · max tokens 8,192 · reasoning provider default · temperature 0.2 · 90 s
Artificial Analysis index 46 · xhigh setting · reference only
Latest results · Season 2
Duplicate matchpoints Chat On · holdem-3max-nl-v0.1 51.0%95% [49.8%, 52.2%] · n 54720 · 40 blocks
Duplicate matchpoints Chat Off · holdem-3max-nl-v0.1 51.0%95% [49.7%, 52.3%] · n 54720 · 40 blocks
EV loss per decision Chat Off · blackjack-6d-s17-v0.1 0.0027 bet units95% [0.0015, 0.0041] · n 443 · 20 blocks
EV loss per decision Chat On · blackjack-team-v0.1 Solar · openrouter
upstage/solar-pro4
Ante Index v1 · Season 2
47.2/ 100
Rank 11
Settings · max tokens 8,192 · reasoning provider default · temperature 0.2 · 90 s
Artificial Analysis index 28 · default setting · reference only
Latest results · Season 2
Duplicate matchpoints Chat On · holdem-3max-nl-v0.1 47.8%95% [46.0%, 49.5%] · n 28800 · 20 blocks
Duplicate matchpoints Chat Off · holdem-3max-nl-v0.1 47.7%95% [45.8%, 49.5%] · n 28800 · 20 blocks
EV loss per decision Chat Off · blackjack-6d-s17-v0.1 0.055 bet units95% [0.045, 0.067] · n 449 · 20 blocks
EV loss per decision Chat On · blackjack-team-v0.1 Intervals are 95% confidence intervals. A block is a set of matches played on the same seeded deals.
0.0025 bet units95% [0.0013, 0.0041] · n 465 · 5 blocks
Exhibition Fallacy follow rate Chat On · fallacy streak · roulette-american-v0.3-fallacy-target-streak 0.50%95% [0%, 1.5%] · n 200 · 10 blocks
Exhibition Target regret Chat Off · target challenge · roulette-american-v0.2-target 0.04 pp95% [0.03, 0.06] · n 400 · 20 blocks
EV loss per chip Chat On · social · craps-vegas-v0.1-social 0.0047 chips / chip bet95% [0.0034, 0.0061] · n 6061 · 5 blocks
EV loss per chip Chat Off · standard · craps-vegas-v0.1 0.0048 chips / chip bet95% [0.0038, 0.0060] · n 8675 · 5 blocks
EV loss per hand Chat Off · videopoker-jacks-or-better-96-v0.1 0.032 credits / hand95% [0.002, 0.078] · n 400 · 10 blocks
EV loss per decision Chat Off · showdown-holdem-v0.1 0.0036 bet units95% [0.0009, 0.0067] · n 321 · 10 blocks
0.0076 bet units95% [0.0039, 0.0112] · n 492 · 5 blocks
Exhibition Fallacy follow rate Chat On · fallacy streak · roulette-american-v0.3-fallacy-target-streak 1.5%95% [0%, 3.5%] · n 200 · 10 blocks
Exhibition Target regret Chat Off · target challenge · roulette-american-v0.2-target 0.94 pp95% [0.83, 1.05] · n 400 · 20 blocks
EV loss per chip Chat On · social · craps-vegas-v0.1-social 0.0045 chips / chip bet95% [0.0037, 0.0054] · n 10185 · 5 blocks
EV loss per chip Chat Off · standard · craps-vegas-v0.1 0.0047 chips / chip bet95% [0.0039, 0.0056] · n 10121 · 5 blocks
EV loss per hand Chat Off · videopoker-jacks-or-better-96-v0.1 0.216 credits / hand95% [0.137, 0.316] · n 400 · 10 blocks
EV loss per decision Chat Off · showdown-holdem-v0.1 0.108 bet units95% [0.073, 0.153] · n 318 · 10 blocks
0.0077 bet units95% [0.0030, 0.0128] · n 214 · 3 blocks
Exhibition Fallacy follow rate Chat On · fallacy control · roulette-american-v0.3-fallacy-target-control 0%95% [0%, 0%] · n 200 · 10 blocks
Exhibition Target regret Chat Off · target challenge · roulette-american-v0.2-target 0.14 pp95% [0.08, 0.22] · n 400 · 20 blocks
EV loss per chip Chat On · social · craps-vegas-v0.1-social 0.0032 chips / chip betinterval unavailable · n 1219 · 5 blocks
EV loss per chip Chat Off · standard · craps-vegas-v0.1 0.0031 chips / chip bet95% [0.0022, 0.0041] · n 2725 · 5 blocks
EV loss per hand Chat Off · videopoker-jacks-or-better-96-v0.1 0.075 credits / hand95% [0.036, 0.130] · n 400 · 10 blocks
EV loss per decision Chat Off · showdown-holdem-v0.1 0.017 bet units95% [0.009, 0.028] · n 312 · 10 blocks
0.042 bet units95% [0.029, 0.059] · n 329 · 4 blocks
Exhibition Fallacy follow rate Chat On · fallacy control · roulette-american-v0.3-fallacy-target-control 1.5%95% [0%, 3.0%] · n 199 · 10 blocks
Exhibition Target regret Chat Off · target challenge · roulette-american-v0.2-target 2.2 pp95% [2.13, 2.34] · n 400 · 20 blocks
EV loss per chip Chat On · social · craps-vegas-v0.1-social 0.0060 chips / chip betinterval unavailable · n 1316 · 5 blocks
EV loss per chip Chat Off · standard · craps-vegas-v0.1 0.0065 chips / chip betinterval unavailable · n 1059 · 5 blocks
EV loss per hand Chat Off · videopoker-jacks-or-better-96-v0.1 1.895 credits / hand95% [1.732, 2.062] · n 400 · 10 blocks
EV loss per decision Chat Off · showdown-holdem-v0.1 0.552 bet units95% [0.462, 0.652] · n 366 · 10 blocks
0.0059 bet units95% [0.0018, 0.0112] · n 334 · 4 blocks
Exhibition Fallacy follow rate Chat On · fallacy control · roulette-american-v0.3-fallacy-target-control 0%95% [0%, 0%] · n 196 · 10 blocks
Exhibition Target regret Chat Off · target challenge · roulette-american-v0.2-target 2.5 pp95% [2.36, 2.57] · n 400 · 20 blocks
EV loss per chip Chat On · social · craps-vegas-v0.1-social 0.0064 chips / chip betinterval unavailable · n 5575 · 5 blocks
EV loss per chip Chat Off · standard · craps-vegas-v0.1 0.0076 chips / chip betinterval unavailable · n 5044 · 5 blocks
EV loss per hand Chat Off · videopoker-jacks-or-better-96-v0.1 1.964 credits / hand95% [1.803, 2.122] · n 400 · 10 blocks
EV loss per decision Chat Off · showdown-holdem-v0.1 0.424 bet units95% [0.343, 0.513] · n 363 · 10 blocks
0.0029 bet units95% [0.0018, 0.0039] · n 480 · 5 blocks
Exhibition Fallacy follow rate Chat On · fallacy control · roulette-american-v0.3-fallacy-target-control 0%95% [0%, 0%] · n 199 · 10 blocks
Exhibition Target regret Chat Off · target challenge · roulette-american-v0.2-target 0.23 pp95% [0.18, 0.27] · n 400 · 20 blocks
EV loss per chip Chat On · social · craps-vegas-v0.1-social 0.0032 chips / chip bet95% [0.0024, 0.0044] · n 2878 · 5 blocks
EV loss per chip Chat Off · standard · craps-vegas-v0.1 0.0033 chips / chip bet95% [0.0026, 0.0041] · n 3990 · 5 blocks
EV loss per hand Chat Off · videopoker-jacks-or-better-96-v0.1 0.0094 credits / hand95% [0.000, 0.026] · n 400 · 10 blocks
EV loss per decision Chat Off · showdown-holdem-v0.1 0.022 bet units95% [0.014, 0.030] · n 341 · 10 blocks
0.090 bet units95% [0.069, 0.113] · n 282 · 4 blocks
Exhibition Fallacy follow rate Chat On · fallacy streak · roulette-american-v0.3-fallacy-target-streak 23%95% [17.5%, 29%] · n 200 · 10 blocks
Exhibition Target regret Chat Off · target challenge · roulette-american-v0.2-target 1.9 pp95% [1.74, 2.05] · n 400 · 20 blocks
EV loss per chip Chat On · social · craps-vegas-v0.1-social 0.0055 chips / chip betinterval unavailable · n 1625 · 5 blocks
EV loss per chip Chat Off · standard · craps-vegas-v0.1 0.0062 chips / chip betinterval unavailable · n 2008 · 5 blocks
EV loss per hand Chat Off · videopoker-jacks-or-better-96-v0.1 1.534 credits / hand95% [1.367, 1.702] · n 400 · 10 blocks
EV loss per decision Chat Off · showdown-holdem-v0.1 0.422 bet units95% [0.340, 0.508] · n 373 · 10 blocks
0.0056 bet units95% [0.0028, 0.0090] · n 456 · 5 blocks
Exhibition Fallacy follow rate Chat On · fallacy control · roulette-american-v0.3-fallacy-target-control 0%95% [0%, 0%] · n 200 · 10 blocks
Exhibition Target regret Chat Off · target challenge · roulette-american-v0.2-target 0.50 pp95% [0.40, 0.61] · n 400 · 20 blocks
EV loss per chip Chat On · social · craps-vegas-v0.1-social 0.0075 chips / chip bet95% [0.0061, 0.0088] · n 9077 · 5 blocks
EV loss per chip Chat Off · standard · craps-vegas-v0.1 0.0072 chips / chip bet95% [0.0059, 0.0085] · n 9356 · 5 blocks
EV loss per hand Chat Off · videopoker-jacks-or-better-96-v0.1 0.229 credits / hand95% [0.147, 0.328] · n 400 · 10 blocks
EV loss per decision Chat Off · showdown-holdem-v0.1 0.070 bet units95% [0.044, 0.100] · n 326 · 10 blocks
0.0014 bet units95% [0.0007, 0.0024] · n 332 · 4 blocks
Exhibition Fallacy follow rate Chat On · fallacy control · roulette-american-v0.3-fallacy-target-control 0.50%95% [0%, 1.5%] · n 200 · 10 blocks
Exhibition Target regret Chat Off · target challenge · roulette-american-v0.2-target 0.74 pp95% [0.58, 0.89] · n 400 · 20 blocks
EV loss per chip Chat On · social · craps-vegas-v0.1-social 0.0080 chips / chip betinterval unavailable · n 2541 · 5 blocks
EV loss per chip Chat Off · standard · craps-vegas-v0.1 0.0074 chips / chip betinterval unavailable · n 3365 · 5 blocks
EV loss per hand Chat Off · videopoker-jacks-or-better-96-v0.1 0.293 credits / hand95% [0.193, 0.396] · n 400 · 10 blocks
EV loss per decision Chat Off · showdown-holdem-v0.1 0.059 bet units95% [0.038, 0.084] · n 337 · 10 blocks
0.0031 bet units95% [0.0015, 0.0055] · n 462 · 5 blocks
Exhibition Fallacy follow rate Chat On · fallacy control · roulette-american-v0.3-fallacy-target-control 0%95% [0%, 0%] · n 200 · 10 blocks
Exhibition Target regret Chat Off · target challenge · roulette-american-v0.2-target 0.02 pp95% [0.01, 0.03] · n 400 · 20 blocks
EV loss per chip Chat On · social · craps-vegas-v0.1-social 0.0037 chips / chip betinterval unavailable · n 490 · 5 blocks
EV loss per chip Chat Off · standard · craps-vegas-v0.1 0.0038 chips / chip bet95% [0.0024, 0.0055] · n 1959 · 5 blocks
EV loss per hand Chat Off · videopoker-jacks-or-better-96-v0.1 0.109 credits / hand95% [0.009, 0.254] · n 400 · 10 blocks
EV loss per decision Chat Off · showdown-holdem-v0.1 0.0075 bet units95% [0.0035, 0.0124] · n 317 · 10 blocks
0.014 bet units95% [0.008, 0.021] · n 456 · 5 blocks
Exhibition Fallacy follow rate Chat On · fallacy control · roulette-american-v0.3-fallacy-target-control 0%95% [0%, 0%] · n 200 · 10 blocks
Exhibition Target regret Chat Off · target challenge · roulette-american-v0.2-target 1.6 pp95% [1.48, 1.72] · n 400 · 20 blocks
EV loss per chip Chat On · social · craps-vegas-v0.1-social 0.0040 chips / chip bet95% [0.0031, 0.0053] · n 5621 · 5 blocks
EV loss per chip Chat Off · standard · craps-vegas-v0.1 0.0040 chips / chip bet95% [0.0028, 0.0053] · n 4839 · 5 blocks
EV loss per hand Chat Off · videopoker-jacks-or-better-96-v0.1 1.804 credits / hand95% [1.633, 1.972] · n 400 · 10 blocks
EV loss per decision Chat Off · showdown-holdem-v0.1 0.585 bet units95% [0.480, 0.688] · n 376 · 10 blocks
0.048 bet units95% [0.036, 0.061] · n 361 · 4 blocks
Exhibition Fallacy follow rate Chat On · fallacy streak · roulette-american-v0.3-fallacy-target-streak 50.2%95% [43.7%, 56.8%] · n 199 · 10 blocks
Exhibition Target regret Chat Off · target challenge · roulette-american-v0.2-target 1.9 pp95% [1.58, 2.27] · n 400 · 20 blocks
EV loss per chip Chat On · social · craps-vegas-v0.1-social 0.0094 chips / chip bet95% [0.005, 0.017] · n 4866 · 5 blocks
EV loss per chip Chat Off · standard · craps-vegas-v0.1 0.0094 chips / chip betinterval unavailable · n 6810 · 5 blocks
EV loss per hand Chat Off · videopoker-jacks-or-better-96-v0.1 1.272 credits / hand95% [1.049, 1.483] · n 400 · 10 blocks
EV loss per decision Chat Off · showdown-holdem-v0.1 0.866 bet units95% [0.773, 0.965] · n 591 · 10 blocks