Model comparison
NVIDIA-Nemotron-3-Nano-4B-BF16 vs Qwen3-VL-8B-Instruct-abliterated-v1
At Q4_K_M, NVIDIA-Nemotron-3-Nano-4B-BF16 is the smaller download — 2,837,072,896 bytes against 5,027,785,792.
From the file· summed bytes, KV per layer
Side by side
| NVIDIA-Nemotron-3-Nano-4B-BF16 | Qwen3-VL-8B-Instruct-abliterated-v1 | |
|---|---|---|
| Parameters | 4.0B | 8.8B |
| Architecture | nemotron_h | qwen3vl |
| Layers | 42 | — |
| Native context | 262,144 | — |
| Mixture of experts | no | no |
| Quantizations published | 42 | 48 |
| Smallest quantization | 2.03 GiB | 1.97 GiB |
| Q4_K_M | 2.64 GiB | 4.68 GiB |
| Licence | other | apache-2.0 |
KV cache by context
the term that decides long-context viability
| Context | NVIDIA-Nemotron-3-Nano-4B-BF16 | Qwen3-VL-8B-Instruct-abliterated-v1 | Ratio |
|---|---|---|---|
| 4,096 | 0.66 GiB | — | — |
| 8,192 | 1.31 GiB | — | — |
| 16,384 | 2.63 GiB | — | — |
| 32,768 | 5.25 GiB | — | — |
| 65,536 | 10.50 GiB | — | — |
| 131,072 | 21.00 GiB | — | — |