DeepSeek·released 2023-11-021 source
DeepSeek Coder 1.3B benchmark scores: 4 benchmarks tracked. 0% of its scores are independently reproduced — the rest are self-reported or unverified.
Leads on 0 of 4 benchmarks·best result 6.6 on Winogrande (self-reported)·0 of 4 independently reproduced
| Benchmark | Score | Evidence status | Measured |
|---|---|---|---|
| Winogrande | 6.6accuracy (%) | self-reported· optimizedT1 | 2023-11-02Qwen2.5-Coder Technical Report |
| GSM8K | 4.4accuracy (%) | self-reported· optimizedT1 | 2023-11-02Qwen2.5-Coder Technical Report |
| MMLU | 1.07accuracy (%) | self-reported· optimizedT1 | 2023-11-02Qwen2.5-Coder Technical Report |
| ARC AI2 | 0.53accuracy (%) | self-reported· optimizedT1 | 2023-11-02Qwen2.5-Coder Technical Report |
Claims drawn from cited facts, not live model generation.
DeepSeek introduced this model in 2023, a compact-scale 1.3-billion-parameter design. That scale favors efficiency and lower deployment cost, though it offers less headroom than larger models.
4 cited facts
This model is tracked on 4 benchmarks and currently holds no top score on any of them. Since these numbers are self-reported, they should be read as vendor claims rather than independently verified results.
3 cited facts
Relative to the leader under the disclosed harness, this model trails by an average of 83.91 points, which is a wide gap and places it well behind the front-of-pack. It holds the top score on none of the benchmark leaderboards, so it shows no category-leading result yet.
2 cited facts
2 facts cross-checked across data sources: 2 single-source.
Source facts, citations, and refresh stamp for this record.
Sources: curated_param_counts, epoch_benchmarks
DeepSeek Coder 1.3B is an AI model developed by DeepSeek, released 2023-11-02. Its benchmark record below tags every score as independently reproduced, self-reported, or unverified.
DeepSeek Coder 1.3B has recorded scores on 4 benchmarks, each shown with its evidence status.
DeepSeek Coder 1.3B has recorded scores on 4 benchmarks — Winogrande, MMLU, ARC AI2, GSM8K. The full table above shows each score with its evidence status.
0 of 4 recorded scores (0%) are independently reproduced rather than self-reported by the lab.
DeepSeek Coder 1.3B has 4 tracked claims: 4 self-reported.