Deepseek Coder 1.3B Base — Benchmarks

Benchmark scores for Deepseek Coder 1.3B Base aggregated from public leaderboards, with how it ranks among open models. See hardware requirements for what you need to run it.

Overall rank: #122 of 123 open modelscomposite 8.1/100 across 4 benchmarks in 3 categories · methodology

Knowledge

BenchmarkScoreOpen rankAll models
ARC Challenge25.4%#50 / 51#75 / 77
MMLU25.8%#76 / 77#134 / 136

Math

BenchmarkScoreOpen rankAll models
GSM8K4.4%#60 / 60#85 / 93

Reasoning

BenchmarkScoreOpen rankAll models
WinoGrande53.3%#45 / 46#79 / 80

Scores aggregated from public benchmark sources (each linked from the benchmark pages). llmrun does not run these benchmarks.