Cheese Dossier · CF-1.3
GPT-4.1 nano
openai · Cheese #169· 5 signals · specialty gpqa · Proprietary
CHEESE INDEX
31.7
DEPTH
44.5
BALANCE
89.5
CONSENSUS
56.9
STABILITY
78.6
COVERAGE
31%
VALUE
63.32
I / O
$0.1 / $0.4
SPECIALTY
gpqa +-23.8
BEST FIT
budget 52.6
深度雷達
實線＝本模型 · 虛線＝群體中位
芝士軸
coding
horizon
Cheese Index = 可用軸中位數；Balance = 三軸分歧小。
能力細節
gpqa
math
knowledge
instruction
部署建議
中段指數：適合成本敏感路徑。優先看 Pareto／Value；用深度雷達找專長匹配，再與更高指數模型 A/B。
鄰近差分
vs MAI-Thinking-1
coding-26.1
agent—
horizon—
指數鄰近
- MAI-Thinking-1 31.0
- o3-mini 32.4
- U2 30.8
- Seed 2.1 Turbo 32.8
- Qwen3.5 397B 33.0
同組織
- GPT-5.6 Sol 72.7
- GPT-5.6 Terra 69.8
- GPT-5.6 Luna 67.2
- GPT-5.5 63.7