Benchgen

EnterpriseClawBench — Results

RankModelScore
1gpt-5-566.3
2claude-sonnet-4-664.4
3claude-opus-4-663.5
4claude-sonnet-4-663.2
5gpt-5-562.9
6claude-sonnet-4-662.3
7gpt-5-562
8gpt-5-561.4
9claude-opus-4-660.2
10deepseek-v4-pro-max58.8
11deepseek-v4-pro-max58.6
12claude-opus-4-657.3
13deepseek-v4-pro-max56.2
14kimi-k2-656.1
15minimax-m352.5
16minimax-m352
17kimi-k2-651.9
18kimi-k2-651.5
19claude-haiku-4-551.3
20claude-haiku-4-550.7
21claude-sonnet-4-645.8
22claude-haiku-4-542.9
23claude-opus-4-642.5
24minimax-m342.1
25claude-haiku-4-540.5

EnterpriseClawBench — Phases

1 phaseActive

Phase 1·

EnterpriseClawBench

Jun 23, 2026
Current