DeepSeek-V4-Flash-0731 vs DeepSeek-R1-0528-NVFP4-v2
Specs, VRAM requirements and download trends — updated 5 September 2026.
deepseek-ai · A · Open weights
Params
304.2 B
Context
1,048,576
30d
2.7 M
mit · commercial OK
nvidia · B · Open weights
Params
393.6 B
Context
163,840
30d
190 K
mit · commercial OK
Specification comparison
Differences are highlighted; identical values are muted.
| Specification | DeepSeek-V4-Flash-0731 | DeepSeek-R1-0528-NVFP4-v2 |
|---|---|---|
| Deployment | Open weights · also via API | Open weights |
| Parameters | 304.2 B | 393.6 B |
| Architecture | DeepseekV4ForCausalLM | DeepseekV3ForCausalLM |
| Context length | 1,048,576 | 163,840 |
| Licence | mit | mit |
| Commercial use | Allowed | Allowed |
| Languages | — | — |
| Downloads 30d | 2,664,769 | 189,945 |
| Downloads all time | 13.2 M | 5.3 M |
| Quantizations on the Hub | SAFETENSORS | SAFETENSORS |
| Gated | No | No |
| First seen on the Hub | 2026-07-31 | 2025-07-21 |
| API input price / 1M tokens | $0.043 | not on OpenRouter |
| API output price / 1M tokens | $0.30 | not on OpenRouter |
| Est. cost per 1M input + 1M output tokens | $0.34 | not on OpenRouter |
| Cached input / 1M tokens | $0.035 | not on OpenRouter |
| AA Intelligence Index | — | not on OpenRouter |
| AA Coding Index | 69.1 | not on OpenRouter |
| AA Agentic Index | 41 | not on OpenRouter |
| Max output tokens | 943,718 | not on OpenRouter |
| API providers | 24 | not on OpenRouter |
| Tool calling | Yes | not on OpenRouter |
| Reasoning mode | Yes | not on OpenRouter |
API cost
Estimated cost of 1M input + 1M output tokens via OpenRouter
DeepSeek-V4-Flash-0731
$0.34
Download trend
Daily snapshots, last 90 days (14 Aug – 11 Nov)
4.7 M190 K
DeepSeek-V4-Flash-0731
DeepSeek-R1-0528-NVFP4-v2
VRAM side by side
On RTX 4090 · 24 GB · 8K context unless noted
DeepSeek-R1-0528-NVFP4-v2 · fp16 @ 160K ctx
1,636.1 GB / 24 GB
DeepSeek-V4-Flash-0731 · fp16 @ 1024K ctx
6,024.3 GB / 24 GB