Model comparison
Qwen3.6-27B-Omnimerge-v4-MTP-GGUF vs Qwen3-30B-A3B-Thinking-2507
These two publish different quantization sets; the table below has the exact sizes.
From the file· summed bytes, KV per layer
Side by side
| Qwen3.6-27B-Omnimerge-v4-MTP-GGUF | Qwen3-30B-A3B-Thinking-2507 | |
|---|---|---|
| Parameters | 27.3B | 30.5B |
| Architecture | qwen35 | qwen3moe |
| Layers | — | 48 |
| Native context | — | 262,144 |
| Mixture of experts | no | yes, 128 experts |
| Quantizations published | 1 | 51 |
| Smallest quantization | 10.64 GiB | 7.05 GiB |
| Q4_K_M | — | 17.28 GiB |
| Licence | other | apache-2.0 |
KV cache by context
the term that decides long-context viability
| Context | Qwen3.6-27B-Omnimerge-v4-MTP-GGUF | Qwen3-30B-A3B-Thinking-2507 | Ratio |
|---|---|---|---|
| 4,096 | — | 0.38 GiB | — |
| 8,192 | — | 0.75 GiB | — |
| 16,384 | — | 1.50 GiB | — |
| 32,768 | — | 3.00 GiB | — |
| 65,536 | — | 6.00 GiB | — |
| 131,072 | — | 12.00 GiB | — |