Model comparison
deepseek-coder-7b-instruct-v1.5 vs Qwen3-VL-8B-Instruct-abliterated-v1
At Q4_K_M, deepseek-coder-7b-instruct-v1.5 is the smaller download — 4,223,359,872 bytes against 5,027,785,792.
From the file· summed bytes, KV per layer
Side by side
| deepseek-coder-7b-instruct-v1.5 | Qwen3-VL-8B-Instruct-abliterated-v1 | |
|---|---|---|
| Parameters | 6.9B | 8.8B |
| Architecture | llama | qwen3vl |
| Layers | 30 | — |
| Native context | 4,096 | — |
| Mixture of experts | no | no |
| Quantizations published | 49 | 48 |
| Smallest quantization | 1.61 GiB | 1.97 GiB |
| Q4_K_M | 3.93 GiB | 4.68 GiB |
| Licence | other | apache-2.0 |
KV cache by context
the term that decides long-context viability
| Context | deepseek-coder-7b-instruct-v1.5 | Qwen3-VL-8B-Instruct-abliterated-v1 | Ratio |
|---|---|---|---|
| 4,096 | 1.88 GiB | — | — |
| 8,192 | 3.75 GiB | — | — |
| 16,384 | 7.50 GiB | — | — |
| 32,768 | 15.00 GiB | — | — |
| 65,536 | 30.00 GiB | — | — |
| 131,072 | 60.00 GiB | — | — |