DeepSeek-V4-Flash vs NVIDIA-Nemotron-3-Super-120B-A12B-FP8
Specs, VRAM requirements and download trends — updated 6 August 2026.
Specification comparison
Differences are highlighted; identical values are muted.
| Specification | DeepSeek-V4-Flash | NVIDIA-Nemotron-3-Super-120B-A12B-FP8 |
|---|---|---|
| Parameters | 158.1 B | 123.6 B |
| Architecture | DeepseekV4ForCausalLM | NemotronHForCausalLM |
| Context length | 1,048,576 | 262,144 |
| Licence | mit | other |
| Commercial use | Allowed | Unknown |
| Languages | — | 7 |
| Downloads 30d | 2,737,621 | 208,471 |
| Downloads all time | 9.4 M | 2.4 M |
| Quantizations on the Hub | SAFETENSORS | SAFETENSORS |
| Gated | No | No |
| First seen on the Hub | 2026-04-22 | 2026-03-10 |
Download trend
Daily snapshots, last 10 days (28 Jul – 6 Aug)
3.1 M208 K
DeepSeek-V4-Flash
NVIDIA-Nemotron-3-Super-120B-A12B-FP8
VRAM side by side
On RTX 4090 · 24 GB · 8K context unless noted
NVIDIA-Nemotron-3-Super-120B-A12B-FP8 · fp16 @ 256K ctx
735.0 GB / 24 GB
DeepSeek-V4-Flash · fp16 @ 1024K ctx
3,211.0 GB / 24 GB