Model

DeepSeek-Coder-Instruct 1.3B

DeepSeek · open weights

Knowledge & Reasoning

Benchmark	Score	Metric	Harness	Source	Confidence	Link
MMLU	25.8	exact_match	N/A	Independent	medium	arxiv.org
WinoGrande	53.3	accuracy	N/A	Independent	medium	arxiv.org

Benchmark	Score	Metric	Harness	Source	Confidence	Link
GSM8K	4.4	exact_match	N/A	Independent	medium	arxiv.org

Benchmark	Score	Metric	Harness	Source	Confidence	Link
MBPP	49.4	accuracy	N/A	Independent	medium	arxiv.org
SAFIM	52.6	average	N/A	Independent	medium	arxiv.org
SCCGBench	0.0	mock	N/A	Independent	medium	arxiv.org