Connections Eval — One-Shot Box Score

Latest one-shot runs for 55 models (20 games each, one submission per game, max 100 pts; sorted by points, avg time, cost) · Classic (multi-turn) leaderboard →
ModelDatePTSPTS%WGRPTRAPINVAVG/GTOK/GCOST$/G
openai/gpt-6-astra2026-09-049494.0%2080/8034015s1.1k$0.67$0.034
x-ai/grok-4.62026-08-129191.0%1978/8032027s3.1k$0.31$0.015
meta/muse-spark-1.12026-07-229090.0%2080/8030038s5.4k$0.42$0.021
google/gemini-3.8-flash2026-09-028888.0%2080/802808s2.5k$0.15$0.008
google/gemini-3.1-pro-preview2026-07-228686.0%2080/8026014s2.2k$0.42$0.021
openai/gpt-5.6-terra2026-07-228686.0%2080/8026019s2.0k$0.47$0.024
anthropic/claude-fable-52026-07-228585.0%1978/8026033s1.7k$1.08$0.054
google/gemini-3-flash-preview2026-07-228484.0%2080/8024010s2.6k$0.13$0.006
meta/muse-spark-1.32026-09-028484.0%1977/8026015s2.6k$0.19$0.009
openai/gpt-5.6-sol-pro2026-07-228484.0%2080/8024047s10.0k$2.68$0.134
moonshotai/kimi-k32026-07-228484.0%1874/8028051s2.3k$0.55$0.027
sakana/fugu-ultra2026-07-228484.0%2080/802402m58s24.1k$9.54$0.477
openai/gpt-5.6-sol2026-07-228383.0%1772/8028127s1.5k$0.63$0.032
anthropic/claude-fable-5.12026-09-018282.0%1977/8024018s1.9k$1.32$0.066
x-ai/grok-4.202026-07-228282.0%2080/802201m20s7.9k$0.37$0.019
google/gemini-3.7-flash2026-08-138080.0%2080/802004s1.7k$0.10$0.005
meta/muse-spark-1.3-contributor2026-09-028080.0%1773/8024051s2.8k$0.01$0.001
deepseek/deepseek-v4-pro-08132026-08-128080.0%2080/802009m20s16.1k$0.28$0.014
z-ai/glm-5.3-flash2026-08-287878.0%2080/801801m52s4.1k$0.02$0.001
meta/muse-spark-1.22026-08-077777.0%1875/8020016s3.0k$0.22$0.011
openai/gpt-5.6-terra-pro2026-07-227777.0%1978/8018036s13.4k$1.98$0.099
deepseek/deepseek-v4-flash-07312026-07-317777.0%1976/802012m50s20.4k$0.11$0.006
x-ai/grok-4.52026-07-227676.0%1874/802011m3s4.4k$0.44$0.022
qwen/qwen3.8-flash2026-09-017575.0%1976/8018113m35s59.6k$0.56$0.028
anthropic/claude-opus-52026-07-247474.0%2080/801408s1.4k$0.40$0.020
openai/gpt-5.6-luna2026-07-307474.0%1464/8024116s2.1k$0.02$0.001
meta/muse-spark-1.2-contributor2026-08-287474.0%1977/8016030s3.0k$0.01$0.001
z-ai/glm-5v-turbo2026-07-227474.0%1670/802001m6s5.8k$0.44$0.022
bytedance-seed/seed-2-1-turbo2026-08-127373.0%1976/801607m0s26.0k$1.28$0.064
openai/gpt-5.6-luna-pro2026-07-307272.0%1773/8016028s12.4k$0.07$0.004
anthropic/claude-sonnet-52026-07-227171.0%1671/8016015s2.1k$0.30$0.015
anthropic/claude-opus-4.52026-07-227070.0%1874/8014056s4.1k$1.82$0.091
google/gemini-3.6-flash2026-07-226969.0%1976/8012011s2.8k$0.36$0.018
deepseek/deepseek-v4-flash2026-07-226871.6%1464/801817m12s15.0k$0.08$0.004
google/gemini-3.5-flash-lite2026-07-226767.0%1669/801408s2.8k$0.12$0.006
google/gemini-3.5-flash2026-07-226767.0%1768/8016310s2.7k$0.42$0.021
thinkingmachines/inkling2026-07-226767.0%1667/801612m42s3.2k$0.22$0.011
qwen/qwen3.8-max2026-08-126767.0%1667/801622m56s10.9k$1.27$0.064
qwen/qwen3.8-2.4t-a95b2026-08-126666.0%1462/8018312m1s16.8k$1.61$0.080
anthropic/claude-opus-4.72026-07-226565.0%1976/808114s1.7k$0.49$0.024
stepfun/step-3.7-flash2026-07-226363.0%1261/801411m23s14.3k$0.32$0.016
meta/muse-glimmer-30b2026-08-126161.0%1053/8018127s4.1k$0.09$0.005
anthropic/claude-haiku-4.52026-07-226060.0%1359/8014130s4.3k$0.39$0.019
z-ai/glm-5.32026-08-285757.0%1768/80635m30s16.4k$1.41$0.070
bytedance-seed/seed-2.0-code2026-08-125656.0%1357/801201m58s7.9k$0.44$0.022
tencent/hy4-preview2026-08-285656.0%1460/801005m39s18.1k$0.89$0.045
qwen/qwen3.8-27b2026-09-025454.0%1054/801014m20s19.4k$0.76$0.038
inception/mercury-2.5-preview2026-09-014545.0%847/80608s5.9k$0.01$0.001
inception/mercury-22026-07-223939.0%429/801434s2.6k$0.03$0.002
qwen/qwen3-max-thinking2026-07-233535.0%1043/802715m58s41.7k$2.49$0.125
ibm-granite/granite-4.2-8b2026-09-012525.0%011/8014111s2.7k$0.01$0.000
nvidia/nemotron-3.5-lightning2026-09-012020.0%216/806117m51s60.6k$0.18$0.009
poolside/laguna-xs-2.12026-07-221717.0%217/80251m59s19.4k$0.05$0.002
upstage/solar-pro42026-09-011616.0%317/802133m34s41.9k$0.05$0.003
ibm-granite/granite-4.1-8b2026-09-021313.0%09/80418s1.7k$0$0.000