Qwen2.5 7B Instruct

Qwen · released 2024-10-16 · apache-2.0 license

Qwen2.5-7B-Instruct is the popular small Qwen2.5 model (Apache-2.0): vendor GPQA 36.4, MMLU-Pro 56.3, LiveCodeBench 28.7, outperforming Gemma2-9B and Llama3.1-8B on most vendor benchmarks.

Key specs

TypeLocal open-weight
Parameters7.62B total
Architectureqwen2
Context window131K tokens
Knowledge cutoff2024-06-30
Modalitiestext
Recommended backends
Minimum viable rig

Benchmark scores

GPQA Diamond36.4%
SWE-bench Verified
AIME
MMLU-Pro56.3%
BFCL v3 (tool use)
Composite score5.36
Community ratingNo reviews yet

VRAM & disk per quantization

QuantVRAMDiskRAMContext
Q4_K_M5.9 GB4.4 GB131K

Strengths & weaknesses

Strengths: Strong small-model math (vendor MATH 75.5); Beats Gemma2-9B and Llama3.1-8B on most vendor evals; Apache-2.0

Weaknesses: IFEval (71.2) below Llama3.1-8B; No native reasoning mode