Model comparison
WizardCoder-Python-13B-V1.0 vs Qwen3-8B
At Q4_K_M, Qwen3-8B is the smaller download — 5,027,783,488 bytes against 7,865,963,424. At long context the gap widens: Qwen3-8B's KV cache at 32K is 5.6× smaller, which usually matters more than the difference in weights.
From the file· summed bytes, KV per layer
Side by side
| WizardCoder-Python-13B-V1.0 | Qwen3-8B | |
|---|---|---|
| Parameters | 13.0B | 8.2B |
| Architecture | llama | qwen3 |
| Layers | 40 | 36 |
| Native context | 16,384 | 40,960 |
| Mixture of experts | no | no |
| Quantizations published | 47 | 51 |
| Smallest quantization | 2.70 GiB | 2.12 GiB |
| Q4_K_M | 7.33 GiB | 4.68 GiB |
| Licence | llama2 | apache-2.0 |
KV cache by context
the term that decides long-context viability
| Context | WizardCoder-Python-13B-V1.0 | Qwen3-8B | Ratio |
|---|---|---|---|
| 4,096 | 3.13 GiB | 0.56 GiB | 5.56× |
| 8,192 | 6.25 GiB | 1.13 GiB | 5.56× |
| 16,384 | 12.50 GiB | 2.25 GiB | 5.56× |
| 32,768 | 25.00 GiB | 4.50 GiB | 5.56× |
| 65,536 | 50.00 GiB | 9.00 GiB | 5.56× |
| 131,072 | 100.00 GiB | 18.00 GiB | 5.56× |