LLM Benchmark Evaluation
Source-mixed reference results; not a controlled head-to-head comparison
Agnes-3.0-Flash
Qwen3.6-35B-A3B
Kimi K2.5
Muse Glimmer
Qwen3.5-27B
DeepSeek V4 Flash
Qwen3.8-27B
Gemini 3.5 Flash
Qwen3.8 Flash Next
MiniMax M3
85.05
84.1
78.9
83.5
85.8
90.8
90.5
92.2
92.3
92.9
GPQA Diamond
68.33
66.7
59
80
72.3
79.7
82
81
79.7
74
AA-LCR
74.2
64.4
43.7
77
75.6
75.8
79.5
76.3
81.3
82.9
IFBench
38.08
35.8
39.6
43.6
39.5
50.3
46.6
53.1
50.6
45.4
SciCode
23
18.8
22.9
27
20.7
40.4
18.4
51.4
24.5
16.7
AA-Omniscience Accuracy