96.4
The large value shows the leading mean for this attempt. Rankings are calculated across repeated attempts rather than from this result alone.
- Attempt
- wave5-5304
- Model
- anthropic/claude-opus-4.7
- Questions asked
- 177
- Seed
- 5304
- Fingerprint
- 77515bc0963aa70a298df419
- When
- 2026-09-18T16:37:41.980434+00:00
- Duration
- 24.2 min
- Cost
- 38.63 USD billed
- Tokens
- 2.25M prompt · 2.18M completion · 4.43M total
z-ai/glm-5.3
mean on this sampled test
· 0.544 USD billed
· 25.0k / 129.1k / 154.1k tok
· 5.5 s / question
80.8
x-ai/grok-4.3
mean on this sampled test
· 0.310 USD billed
· 56.7k / 110.1k / 166.8k tok
· 5.5 s / question
94.7
x-ai/grok-4.5
mean on this sampled test
· 1.20 USD billed
· 59.7k / 184.4k / 244.1k tok
· 9.8 s / question
95.2
qwen/qwen3.7-max
mean on this sampled test
· 0.874 USD billed
· 26.3k / 188.9k / 215.1k tok
· 12.1 s / question
91.9
qwen/qwen3.7-plus
mean on this sampled test
· 0.249 USD billed
· 26.3k / 187.6k / 213.9k tok
· 13.3 s / question
91.7
openai/gpt-5.5-pro
mean on this sampled test
· 19.27 USD billed
· 23.6k / 103.1k / 126.8k tok
· 13.1 s / question
80.2
meta/muse-spark-1.3
mean on this sampled test
· 0.946 USD billed
· 26.9k / 214.6k / 241.5k tok
· 10.7 s / question
69.9
openai/gpt-5.6-sol-pro
mean on this sampled test
· 2.19 USD billed
· 461.1k / 126.9k / 588.0k tok
· 6.7 s / question
96.0
openai/gpt-6-astra-pro
mean on this sampled test
· 6.60 USD billed
· 373.4k / 57.3k / 430.7k tok
· 5.7 s / question
92.9
qwen/qwen3.8-2.4t-a95b
mean on this sampled test
· 1.10 USD billed
· 33.0k / 172.3k / 205.3k tok
· 7.2 s / question
70.1
openai/gpt-5.6-luna-pro
mean on this sampled test
· 0.328 USD billed
· 500.1k / 190.2k / 690.4k tok
· 6.5 s / question
96.2
deepseek/deepseek-v4-pro
mean on this sampled test
· 0.305 USD billed
· 27.2k / 115.6k / 142.8k tok
· 9.9 s / question
85.7
openai/gpt-5.6-terra-pro
mean on this sampled test
· 2.72 USD billed
· 476.9k / 147.5k / 624.3k tok
· 8.0 s / question
94.3
anthropic/claude-opus-4.7
mean on this sampled test
· 0.664 USD billed
· 36.7k / 19.2k / 55.9k tok
· 2.0 s / question
96.4
anthropic/claude-opus-4.8
mean on this sampled test
· 0.929 USD billed
· 35.8k / 30.0k / 65.8k tok
· 2.9 s / question
94.5
deepseek/deepseek-v4.1-flash
mean on this sampled test
· 0.106 USD billed
· 28.0k / 104.3k / 132.3k tok
· 4.8 s / question
77.2
deepseek/deepseek-v4-pro-0813
mean on this sampled test
· 0.303 USD billed
· 31.6k / 103.2k / 134.9k tok
· 7.3 s / question
77.2