Cheese Dossier · CF-1.3

Qwen3.6 Plus

qwen · Cheese #59· 6 signals · specialty gpqa · Proprietary

CHEESE INDEX
53.3
DEPTH
68.6
BALANCE
87.3
CONSENSUS
76.1
STABILITY
99.5
COVERAGE
92%
VALUE
15.23
I / O
$0.5 / $3
SPECIALTY
gpqa +16.3
BEST FIT
long_horizon 59.5

深度雷達

codingagentmathknowleinstrumultimmultilreason

實線=本模型 · 虛線=群體中位

芝士軸

coding40.6
agent53.3
horizon53.0

Cheese Index = 可用軸中位數;Balance = 三軸分歧小。

能力細節

swe ★78.8
gpqa ★90.4
hle28.8
mcp74.1
coding51.9
agent53.3
math ★63.0
knowledge58.3
instruction86.4
multimodal67.5
multilingual ★69.7
reasoning ★85.4

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 Muse Spark 1.1

coding-10.5
agent-14.7
horizon-0.3