List your company
ResearchFormerly WizWand

Reasoning on AIME 2024 (Pass@1)

0.7495Pass@1FP16
0.2213880.3584940.49560.632706May 9, 2026
Updated 5mo ago

Evaluation results

MethodLinks
FP16Model=Qwen3-8B, Speedup=1.0×2026.050.7495
LAQuantModel=Qwen3-8B, Quantization Bit-width=4-bit, Group Size=128, Speedup=2.45×2026.050.7417
ParoQ++Model=Qwen3-8B, Quantization Bit-width=4-bit, Group Size=128, Speedup=2.32×2026.050.7381
ParoQModel=Qwen3-8B, Quantization Bit-width=4-bit, Group Size=128, Speedup=2.32×2026.050.7339
FP16Model=Qwen3-4B, Speedup=1.0×2026.050.7323
GPTQModel=Qwen3-8B, Quantization Bit-width=4-bit, Group Size=128, Speedup=2.45×2026.050.7156
LAQuantModel=Qwen3-4B, Quantization Bit-width=4-bit, Group Size=128, Speedup=2.0×2026.050.7042
ParoQ++Model=Qwen3-4B, Quantization Bit-width=4-bit, Group Size=128, Speedup=1.84×2026.050.7037
R-QATModel=Qwen3-8B, Quantization Bit-width=4-bit, Group Size=128, Speedup=2.45×2026.050.6844
ParoQModel=Qwen3-4B, Quantization Bit-width=4-bit, Group Size=128, Speedup=1.84×2026.050.6729
R-QATModel=Qwen3-4B, Quantization Bit-width=4-bit, Group Size=128, Speedup=2.0×2026.050.6464
GPTQModel=Qwen3-4B, Quantization Bit-width=4-bit, Group Size=128, Speedup=2.0×2026.050.6344
FP16Model=Qwen3-1.7B, Speedup=1.0×2026.050.4584
FP16Model=R1-Distill-Llama-8B, Speedup=1.0×2026.050.4344
LAQuantModel=Qwen3-1.7B, Quantization Bit-width=4-bit, Group Size=128, Speedup=1.64×2026.050.4235
ParoQ++Model=Qwen3-1.7B, Quantization Bit-width=4-bit, Group Size=128, Speedup=1.48×2026.050.4182
LAQuantModel=R1-Distill-Llama-8B, Quantization Bit-width=4-bit, Group Size=128, Speedup=2.46×2026.050.4141
R-QATModel=R1-Distill-Llama-8B, Quantization Bit-width=4-bit, Group Size=128, Speedup=2.46×2026.050.4063
ParoQModel=Qwen3-1.7B, Quantization Bit-width=4-bit, Group Size=128, Speedup=1.48×2026.050.3953
ParoQModel=R1-Distill-Llama-8B, Quantization Bit-width=4-bit, Group Size=128, Speedup=2.33×2026.050.3912
ParoQ++Model=R1-Distill-Llama-8B, Quantization Bit-width=4-bit, Group Size=128, Speedup=2.33×2026.050.3886
GPTQModel=R1-Distill-Llama-8B, Quantization Bit-width=4-bit, Group Size=128, Speedup=2.46×2026.050.3844
R-QATModel=Qwen3-1.7B, Quantization Bit-width=4-bit, Group Size=128, Speedup=1.64×2026.050.3261
GPTQModel=Qwen3-1.7B, Quantization Bit-width=4-bit, Group Size=128, Speedup=1.64×2026.050.2417