QwQ Preview

Preview

Alibaba

•

Open Source

•

32B

# 33

Released

Nov 28, 2024

# 4

Knowledge Cutoff

Sep 24

# 8

Context Length

32K

Benchmarks

# 48

Code LiveBench

37.2%

# 16

LiveCodeBench v5

44.0%

# 20

GPQA Diamond

65.2%

# 21

Reason LiveBench

57.7%

# 18

AIME 2025 I & II

33.5%

# 27

Math LiveBench

50.8%

# 12

MATH 500

90.6%

# 16

MMLU Pro

75.0%

# 33

Halluc. Hughes

12.9%

# 25

AIME 2024

44.0%

# 56

IF LiveBench

35.6%

# 55

Data LiveBench

3.2%

# 53

Lang LiveBench

19.6%

# 55

Avg LiveBench

28.5%