Model comparison
Qwen3-7B-Instruct vs Qwen3.8-27B
At Q4_K_M, Qwen3-7B-Instruct is the smaller download — 4,683,074,272 bytes against 16,810,714,336.
From the file· summed bytes, KV per layer
Side by side
| Qwen3-7B-Instruct | Qwen3.8-27B | |
|---|---|---|
| Parameters | 7.6B | 27.8B |
| Architecture | qwen2 | qwen35 |
| Layers | 28 | 64 |
| Native context | 32,768 | 262,144 |
| Mixture of experts | no | no |
| Quantizations published | 1 | 22 |
| Smallest quantization | 4.36 GiB | 8.39 GiB |
| Q4_K_M | 4.36 GiB | 15.66 GiB |
| Licence | — | apache-2.0 |
KV cache by context
the term that decides long-context viability
| Context | Qwen3-7B-Instruct | Qwen3.8-27B | Ratio |
|---|---|---|---|
| 4,096 | 0.22 GiB | 0.25 GiB | 1.14× |
| 8,192 | 0.44 GiB | 0.50 GiB | 1.14× |
| 16,384 | 0.88 GiB | 1.00 GiB | 1.14× |
| 32,768 | 1.75 GiB | 2.00 GiB | 1.14× |
| 65,536 | 3.50 GiB | 4.00 GiB | 1.14× |
| 131,072 | 7.00 GiB | 8.00 GiB | 1.14× |