Model comparison
GLM-Z1-32B-0414-uncensored-heretic-v2 vs Qwen3.8-27B
At Q4_K_M, Qwen3.8-27B is the smaller download — 16,810,714,336 bytes against 19,686,117,248.
From the file· summed bytes, KV per layer
Side by side
| GLM-Z1-32B-0414-uncensored-heretic-v2 | Qwen3.8-27B | |
|---|---|---|
| Parameters | 32.6B | 27.8B |
| Architecture | glm4 | qwen35 |
| Layers | 61 | 64 |
| Native context | 32,768 | 262,144 |
| Mixture of experts | no | no |
| Quantizations published | 9 | 22 |
| Smallest quantization | 14.80 GiB | 8.39 GiB |
| Q4_K_M | 18.33 GiB | 15.66 GiB |
| Licence | mit | apache-2.0 |
KV cache by context
the term that decides long-context viability
| Context | GLM-Z1-32B-0414-uncensored-heretic-v2 | Qwen3.8-27B | Ratio |
|---|---|---|---|
| 4,096 | 0.24 GiB | 0.25 GiB | 1.05× |
| 8,192 | 0.48 GiB | 0.50 GiB | 1.05× |
| 16,384 | 0.95 GiB | 1.00 GiB | 1.05× |
| 32,768 | 1.91 GiB | 2.00 GiB | 1.05× |
| 65,536 | 3.81 GiB | 4.00 GiB | 1.05× |
| 131,072 | 7.63 GiB | 8.00 GiB | 1.05× |