all versions
archived snapshot
betarebuild: openai/gpt-5.2, openai/gpt-5.6-terra, openai/gpt-5.3-chat, openai/gpt-5
1.12.0__20260711-104416
A frozen leaderboard from the run of July 11, 2026. Read-only history. per-model replays are only kept for the live version.
bench v1.12.0rubric v1.5.036 models5 seeds34 survived · 2 diedjudge anthropic/claude-opus-4-8
Leaderboard · July 11, 2026
HP left · highest first1
openai/gpt-5.2openai
survivedresistedhonestrecalled
96±2.6
$0.0354 · 2,712 HP/$
2
openai/gpt-5.6-terraopenai
survivedresistedhonestrecalled
94±4.2
$0.0322 · 2,919 HP/$
3
openai/gpt-5.3-chatopenai
survivedresistedhonestrecalled
93±3.1
$0.0345 · 2,696 HP/$
4
openai/gpt-5.5openai
survivedresistedunsupportedrecalled
93±1.5
$0.0355 · 2,620 HP/$
5
openai/gpt-5.6-solopenai
survivedresistedunsupportedrecalled
93±1.5
$0.0329 · 2,827 HP/$
6
openai/gpt-5.6-lunaopenai
survivedresistedunsupportedrecalled
92±4.3
$0.0359 · 2,563 HP/$
7
anthropic/claude-sonnet-5anthropic
survivedresistedunsupportedrecalled
89±12.7
$0.0523 · 1,702 HP/$
8
openai/gpt-5.4openai
survivedresistedhonestrecalled
89±7.3
$0.0463 · 1,922 HP/$
9
openai/gpt-5.2-chatopenai
survivedresistedhonestrecalled
88±9.8
$0.0345 · 2,551 HP/$
10
openai/gpt-5.1openai
survivedresistedhonestrecalled
87±3.6
$0.0420 · 2,071 HP/$
11
google/gemini-3-pro-previewgoogle
survivedresistedhonestrecalled
86±11.8
$0.0484 · 1,777 HP/$
12
google/gemini-3.1-pro-previewgoogle
survivedresistedhonestrecalled
84±11
$0.0492 · 1,707 HP/$
13
anthropic/claude-opus-4-8anthropic
survivedresistedhonestrecalled
83±7.2
$0.0471 · 1,762 HP/$
14
anthropic/claude-opus-4-7anthropic
survivedresistedhonestrecalled
78±8.5
$0.0524 · 1,489 HP/$
15
anthropic/claude-opus-4-6anthropic
survivedresistedunsupportedrecalled
77±12.5
$0.0433 · 1,778 HP/$
16
openai/gpt-5.4-miniopenai
survivedresistedhonestrecalled
77±19.8
$0.0489 · 1,575 HP/$
17
anthropic/claude-opus-4-5anthropic
survivedresistedhonestrecalled
75±13
$0.0513 · 1,462 HP/$
18
openai/gpt-5openai
survivedresistedhonestrecalled
67±14.3
$0.0541 · 1,238 HP/$
19
openai/gpt-5-miniopenai
survivedresistedhonestrecalled
66±8.9
$0.0546 · 1,209 HP/$
20
google/gemini-3-flash-previewgoogle
survivedresistedhonestrecalled
65±7
$0.1349 · 482 HP/$
21
openai/gpt-5-nanoopenai
survivedresistedhonestrecalled
65±18.1
$0.0737 · 882 HP/$
22
anthropic/claude-sonnet-4-6anthropic
survivedresistedhonestrecalled
62±3.9
$0.0462 · 1,342 HP/$
23
sference/glm-5.2sference
survivedresistedhonestrecalled
61±2
$0.0472 · 1,292 HP/$
24
anthropic/claude-sonnet-4-5anthropic
survivedresistedhonestrecalled
60±2.8
$0.0494 · 1,215 HP/$
25
alibaba/qwen3.6-plusalibaba
survivedresistedhonestrecalled
57±3.4
$0.0569 · 1,002 HP/$
26
minimaxi/MiniMax-M2.7minimaxi
survivedresistedhonestrecalled
53±7.9
$0.0509 · 1,041 HP/$
27
anthropic/claude-haiku-4-5anthropic
survivedresistedhonestrecalled
47±13.1
$0.0399 · 1,178 HP/$
28
xai/grok-4-fastxai
survivedresistedhallucinatedrecalled
33±9.3
$0.0442 · 747 HP/$
29
groq/openai/gpt-oss-120bgroq
survivedmanipulatederroredrecalled
30±15.7
$0.0481 · 624 HP/$
30
groq/openai/gpt-oss-20bgroq
survivedmanipulatederroredrecalled
19±8.5
$0.0565 · 336 HP/$
31
openai/gpt-4oopenai
survivedmanipulatedhallucinatedrecalled
15±12.6
$0.0261 · 575 HP/$
32
openai/gpt-5-chatopenai
survivedresistedhallucinatedrecalled
14±12.5
$0.0229 · 611 HP/$
33
mistral/mistral-large-latestmistral
survivedresistedhallucinatedrecalled
12±8.5
$0.0311 · 386 HP/$
34
alibaba/qwen-plusalibaba
survivedresistedhallucinatedrecalled
7±10.7
$0.0358 · 196 HP/$
35
anthropic/claude-fable-5anthropic
died @ toolChainresistedrefusedrecalled
8±9.8
$0.0354 · 226 HP/$
36
deepinfra/Qwen/Qwen3-32Bdeepinfra
died @ toolChainmanipulatedunsupportedrecalled
0±0
$0.0544 · 0 HP/$