Back to Calculator Deploy on RunPodDeploy Now
DeepSeek R1 0528
HotUpdated DeepSeek R1 with significantly stronger reasoning — AIME accuracy jumped from 70% to 87.5%, approaching OpenAI o3 and Gemini 2.5 Pro
Specifications
SourceArchitectureTEXT
Parameters671B
Familydeepseek
VRAM (Q4)335.5G
MoE: 37B active.
Minor version upgrade to R1. Now supports system prompts. MIT license. Distilled Qwen3-8B variant also released.
deepseekreasoningmoemathcodingtrending
Run in the Cloud
This model requires enterprise-grade VRAM. Rent GPUs on RunPod and start generating.
Instant Cloud GPUs
Running out of VRAM? Rent a high-end H100 or RTX 4090 on RunPod and deploy in seconds.
Quantization Estimates
| Format | VRAM Need | Tier |
|---|---|---|
| FP16 | 1342.0 GB | Full Precision |
| Q8_0 | 671.0 GB | High |
| Q6_K | 570.4 GB | Excellent |
| Q5_K_M | 469.7 GB | Great |
| Q4_K_M | 335.5 GB | Sweet Spot |
| Q2_K | 201.3 GB | Emergency |
Share this Model
Send these specs directly to your community.