Model comparison
DeepSeek Coder 6.7B vs Qwen2.5-Coder (1.5B)
Neither DeepSeek Coder 6.7B nor Qwen2.5-Coder (1.5B) has enough public benchmark results to be ranked yet; the rows below show what has been published.
Last verified . 5 shared benchmarks.
DeepSeek
Alibaba (Qwen)Summary
- They share 5 benchmarks with published results for both.
Side by side
| DeepSeek Coder 6.7B | Qwen2.5-Coder (1.5B) | |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 37.6 | — |
| Released | 2023-11-02 | 2024-09-18 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 9 | 6 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Not comparable
DeepSeek Coder 6.7B: 35.8 (#219), Qwen2.5-Coder (1.5B): —
| Benchmark | DeepSeek Coder 6.7B | Qwen2.5-Coder (1.5B) |
|---|---|---|
| BigCodeBench Instruct | 35.5% | — |
| BigCodeBench Complete | 43.8% | — |
| HumanEval+ | 71.3% | — |
| MBPP+ | 65.6% | — |
Reasoning Not comparable
DeepSeek Coder 6.7B: —, Qwen2.5-Coder (1.5B): —
| Benchmark | DeepSeek Coder 6.7B | Qwen2.5-Coder (1.5B) |
|---|---|---|
| Epoch Capabilities Index | 89.62 | 113.14 |
| WinoGrande | 57.6% | 72.9% |
| HellaSwag | — | 76.8% |
Math Not comparable
DeepSeek Coder 6.7B: —, Qwen2.5-Coder (1.5B): —
| Benchmark | DeepSeek Coder 6.7B | Qwen2.5-Coder (1.5B) |
|---|---|---|
| GSM8K | 21.3% | 86.7% |
Knowledge Not comparable
DeepSeek Coder 6.7B: —, Qwen2.5-Coder (1.5B): —
| Benchmark | DeepSeek Coder 6.7B | Qwen2.5-Coder (1.5B) |
|---|---|---|
| ARC (AI2) Challenge | 36.4% | 60.9% |
| MMLU | 36.4% | 68% |
Frequently asked questions
Is DeepSeek Coder 6.7B better than Qwen2.5-Coder (1.5B)?
Neither DeepSeek Coder 6.7B nor Qwen2.5-Coder (1.5B) has enough public benchmark results to be ranked yet; the rows below show what has been published.
How many benchmarks do DeepSeek Coder 6.7B and Qwen2.5-Coder (1.5B) share?
5 benchmarks have published results for both models. DeepSeek Coder 6.7B has 9 scored results on Noometry and Qwen2.5-Coder (1.5B) has 6.