List your company

Comprehensive Chinese Evaluation on C-Eval

89AccuracyQwen3
42.77254.773566.77578.7765Aug 13, 2025Oct 1, 2025Nov 19, 2025Jan 7, 2026Feb 25, 2026Apr 15, 2026Jun 3, 2026

Evaluation results

MethodLinks
Qwen3Model=Qwen3, Number of Parameters=32B, Privacy Protocol=Plaintext2026.0389
NTele-R1-32B2025.0888.82
R1-DistillModel=R1-Distill, Number of Parameters=32B, Privacy Protocol=Plaintext2026.0388.34
DS-32B2025.0887.98
AloePriModel=R1-Distill, Number of Parameters=32B, Privacy Protocol=AloePri2026.0387.84
DS-V3.1-Terminus (no_think)Model=DS-V3.1-Terminus (no_think), Number of Parameters=671B, Privacy Protocol=Plaintext2026.0387.67
AloePriModel=Qwen3, Number of Parameters=32B, Privacy Protocol=AloePri2026.0387.64
Qwen3Model=Qwen3, Number of Parameters=14B, Privacy Protocol=Plaintext2026.0387.35
AloePriModel=Qwen3, Number of Parameters=14B, Privacy Protocol=AloePri2026.0387.12
Qwen3-MoE-InstructModel=Qwen3-MoE-Instruct, Number of Parameters=30B-A3B, Privacy Protocol=Plaintext2026.0386.97
AloePriModel=Qwen3-MoE-Instruct, Number of Parameters=30B-A3B, Privacy Protocol=AloePri2026.0386.79
R1-DistillModel=R1-Distill, Number of Parameters=14B, Privacy Protocol=Plaintext2026.0386.18
AloePriModel=R1-Distill, Number of Parameters=14B, Privacy Protocol=AloePri2026.0385.93
AloePriModel=DS-V3.1-Terminus (no_think), Number of Parameters=671B, Privacy Protocol=AloePri2026.0385.65
Qwen2.5 (base)Params=7B, Tokens=18T, Complexity Type=Quadratic, Evaluation Framework=vLLM, Evaluation Method=perplexity-based2025.0981.68
SpikingBrain-76B (76B-A12B)Params=12B/76B, Tokens=+160B, Complexity Type=Hybrid, Evaluation Framework=vLLM, Evaluation Method=perplexity-based2025.0978.89
FPModel=LLADA-1.5, Precision=Full Precision2026.0670.05
FPModel=LLADA, Precision=Full Precision2026.0669.54
MoE + L_ERCNumber of Parameters=15B, Activated Parameters=700M, Loss Function=ERC loss2025.1269
STaR-QuantModel=LLADA-1.5, Quantization=W8A82026.0668.96
STaR-QuantModel=LLADA, Quantization=W8A82026.0667.87
MoENumber of Parameters=15B, Activated Parameters=700M2025.1267.5
FPModel=DREAM, Precision=Full Precision2026.0664.89
STaR-QuantModel=DREAM, Quantization=W8A82026.0662.89
Gemma2-27BParams=27B, Tokens=13T, Complexity Type=Quadratic, Evaluation Framework=vLLM, Evaluation Method=perplexity-based2025.0961.02
Mixtral-8*7B (47B-A13B)Params=13B/47B, Complexity Type=Quadratic, Evaluation Framework=vLLM, Evaluation Method=perplexity-based2025.0954.39
Llama3Model=Llama3, Number of Parameters=8B, Privacy Protocol=Plaintext2026.0350.16
LLama2-70bParams=70B, Tokens=2T, Complexity Type=Quadratic, Evaluation Framework=vLLM, Evaluation Method=perplexity-based2025.0949.26
Jamba (52B-A12B)Params=12B/52B, Complexity Type=Hybrid, Evaluation Framework=vLLM, Evaluation Method=perplexity-based2025.0948.94
AloePriModel=Llama3, Number of Parameters=8B, Privacy Protocol=AloePri2026.0344.55