Share your thoughts, 1 month free Claude Pro on usSee more

Chinese Language Understanding on C-Eval

92.5Accuracy

Kimi-K2

Updated 8d ago

Evaluation Results

Method	Links
Kimi-K2 2026.02		92.5
Qwen2-72B 2024.07		91
DeepSeek V3.2 2026.02		91
DeepSeek V3.1 2026.02		90
Step 3.5 Flash 2026.02		89.6
Qwen1.5-110B 2024.07		89.1
MiMo-V2 Flash 2026.02		87.9
GLM-4.5 2026.02		86.9
Qwen1.5-72B 2024.07		84.1
Qwen2.5 (base) 2025.09		81.6
Qwen3-4B 2025.12		78.5
Qwen2.5-3B 2025.12		74.65
SDAR 8B 2025.12		72.7
NBDiff-7B-INSTRUCT 2025.12		72.5
Qwen2-1.5B 2025.12		71.29
Qwen2-1.5B 2024.07		70.6
SpikingBrain-7B 2025.09		69.8
Qwen2.5-1.5B 2025.12		68.63
Qwen3-1.7B 2025.12		66.7
LLaDA2.0-mini preview 16BA1B 2025.12		66.5
HSA-UL 2025.11		65.98
CAMEL 2026.03		65.3
Llama-3-70B 2024.07		65.2
HySparse 2026.02		65
Full-Attn 2026.02		64.6
LLaDA-MoE 7B-A1B 2025.12		63.9
Human Designed 2026.03		63.7
OPTIMER 2026.03		63.52
DML 2026.03		63.1
SODM 2026.03		62.8
Model-size agnostic 2026.03		62.7
OPTIMER 2026.03		62.56
Qwen1.5-1.8B 2024.07		59.7
Hybrid SWA 2026.02		58.8
HSA-UL 2025.11		58.36
Qwen2-0.5B 2024.07		58.2
Dream-v0 Instruct-7B 2025.12		58
Qwen3-0.6B 2025.12		57.03
TRM-MoE 2025.11		56.87
Mixtral-8x22B 2024.07		54.6
Qwen3 2025.11		54.57
Qwen2.5 2025.11		54.17
YuLan-Mini-2.4B 2025.12		52.32
HySparse 2026.02		52.2
Llama3.1 2025.09		51.46
SEA-LION v4 27B IT 2026.03		50.89
SmolLM3-3B 2025.12		50.84
Full-Attn 2026.02		50.6
Hybrid SWA 2026.02		50.6
Mistral 2025.09		47.04
PCMind-2.1-Kaiyuan-2B 2025.12		46.3
llama-3.2-3B 2025.12		45.67
HSA-UL 2025.11		44.3
Falcon-Mamba 2025.09		41.93
gemma2-2B 2025.12		41.35
NITP 2026.05		40.72
NTP 2026.05		38.97
Gemma 3 27B IT 2026.03		37.96
Zamba-v1 2025.09		36.4
SmolLM2-1.7B 2025.12		35.06
NTP 2026.05		33.53
NITP 2026.05		33.16
NITP 2026.05		33.12
NTP 2026.05		32.31
OLMo-2-0425-1B 2025.12		30.53
llama-3.2-1B 2025.12		29.82
Gemma-2B 2024.07		28
Phi-2 2024.07		23.4