DeepSeek R1 Distill Qwen 32B

deepseek-ai · released 2025-01-20 · mit license

DeepSeek-R1-Distill-Qwen-32B (Jan 2025) distills R1 reasoning into Qwen2.5-32B: AIME24 72.6, MATH-500 94.3, GPQA Diamond 62.1, LiveCodeBench 57.2, beating o1-mini on several. MIT.

Key specs

TypeLocal open-weight
Parameters32.76B total
Architectureqwen2
Context window131K tokens
Knowledge cutoff
Modalitiestext
Recommended backends
Minimum viable rig

Benchmark scores

GPQA Diamond62.1%
SWE-bench Verified
AIME72.6%
MMLU-Pro
BFCL v3 (tool use)
Composite score6.63
Community ratingNo reviews yet

VRAM & disk per quantization

QuantVRAMDiskRAMContext
Q4_K_M20.5 GB19 GB131K

Strengths & weaknesses

Strengths: Strong math reasoning (AIME24 72.6 pass@1, MATH-500 94.3); Beats o1-mini on several benchmarks at dense 32B; MIT model license

Weaknesses: Distilled reasoning; weaker general chat breadth than full R1; Long CoT increases latency/cost