Model comparison
Qwen3.8-27B vs M3.2-24B-Animus-V5-Pro
At Q4_K_M, M3.2-24B-Animus-V5-Pro is the smaller download — 14,333,907,520 bytes against 16,810,714,336.
From the file· summed bytes, KV per layer
Side by side
| Qwen3.8-27B | M3.2-24B-Animus-V5-Pro | |
|---|---|---|
| Parameters | 27.8B | 23.6B |
| Architecture | qwen35 | llama |
| Layers | 64 | — |
| Native context | 262,144 | — |
| Mixture of experts | no | no |
| Quantizations published | 22 | 7 |
| Smallest quantization | 8.39 GiB | 8.28 GiB |
| Q4_K_M | 15.66 GiB | 13.35 GiB |
| Licence | apache-2.0 | mit |
KV cache by context
the term that decides long-context viability
| Context | Qwen3.8-27B | M3.2-24B-Animus-V5-Pro | Ratio |
|---|---|---|---|
| 4,096 | 0.25 GiB | — | — |
| 8,192 | 0.50 GiB | — | — |
| 16,384 | 1.00 GiB | — | — |
| 32,768 | 2.00 GiB | — | — |
| 65,536 | 4.00 GiB | — | — |
| 131,072 | 8.00 GiB | — | — |