DeepSeek-V4-Flash vs DeepSeek-V4-Flash-NVFP4

Specs, VRAM requirements and download trends — updated 6 August 2026.

deepseek-ai · A
Params
158.1 B
Context
1,048,576
30d
2.7 M
mit · commercial OK
nvidia · B
Params
166.7 B
Context
1,048,576
30d
1.1 M
mit · commercial OK

Specification comparison

Differences are highlighted; identical values are muted.

Specification DeepSeek-V4-Flash DeepSeek-V4-Flash-NVFP4
Parameters 158.1 B 166.7 B
Architecture DeepseekV4ForCausalLM DeepseekV4ForCausalLM
Context length 1,048,576 1,048,576
Licence mit mit
Commercial use Allowed Allowed
Languages
Downloads 30d 2,737,621 1,116,514
Downloads all time 9.4 M 1.7 M
Quantizations on the Hub SAFETENSORS SAFETENSORS
Gated No No
First seen on the Hub 2026-04-22 2026-05-18

Download trend

Daily snapshots, last 10 days (28 Jul – 6 Aug)

3.1 M1.1 M
DeepSeek-V4-Flash DeepSeek-V4-Flash-NVFP4

VRAM side by side

On RTX 4090 · 24 GB · 8K context unless noted

DeepSeek-V4-Flash · fp16 @ 1024K ctx 3,211.0 GB / 24 GB
❌ Won’t fit
DeepSeek-V4-Flash-NVFP4 · fp16 @ 1024K ctx 3,386.8 GB / 24 GB
❌ Won’t fit

Adjacent comparisons