DeepSeek on NVIDIA Blackwell

Can I run DeepSeek-R1-Distill-Qwen 32B on an RTX 5070 Ti?

Not at Q4_K_M — it needs 21.5 GB against 14.2 GB available. You would need 2 of these cards.

Won't fit

21.5 GB of 14.2 GB · 151%
022 GB
Weights 18.6 GB
KV cache 2.0 GB
Runtime overhead 0.8 GB
Over the limit 7.2 GB

Short by 7.2 GB. You can run it with 39 of 64 layers on the RTX 5070 Ti and the rest in system RAM, at roughly 4.95 tok/s — usable for batch work, painful for chat. A smaller quantisation or a shorter context is usually the better trade.

Generation4.95tok/s
Prompt processing564tok/s
Max context0tokens
KV per 1K tokens0GB

Every quantisation of DeepSeek-R1-Distill-Qwen 32B on a RTX 5070 Ti

Highlighted row is the highest quality that still fits at 8K context.

QuantisationbpwWeights TotalVerdictMax ctxtok/s
FP16 / BF16 16.00 61.0 GB 63.9 GB Won't fit — 0.81
INT8 / W8A8 8.50 32.1 GB 34.9 GB Won't fit — 1.87
Q8_0 (GGUF) 8.50 32.1 GB 34.9 GB Won't fit — 1.87
FP8 (E4M3) 8.00 30.5 GB 33.3 GB Won't fit — 2.01
Q6_K 6.56 25.0 GB 27.9 GB Won't fit — 2.80
Q5_K_M 5.67 21.6 GB 24.5 GB Won't fit — 3.57
Q5_K_S 5.52 21.1 GB 23.9 GB Won't fit — 3.77
Q4_K_M 4.85 18.6 GB 21.5 GB Won't fit — 4.95
AWQ 4-bit 4.25 18.3 GB 21.2 GB Won't fit — 5.02
GPTQ 4-bit 4.25 18.3 GB 21.2 GB Won't fit — 5.02
MXFP4 4.25 18.3 GB 21.2 GB Won't fit — 5.02
Q4_K_S 4.58 17.6 GB 20.5 GB Won't fit — 5.58
Q4_0 4.55 17.5 GB 20.4 GB Won't fit — 5.61
IQ4_XS 4.25 16.4 GB 19.3 GB Won't fit — 6.68
Q3_K_M 3.91 15.2 GB 18.0 GB Won't fit — 8.55
IQ3_M 3.70 14.4 GB 17.3 GB Won't fit — 9.92
IQ3_XXS 3.06 12.1 GB 15.0 GB Won't fit 5K 24.7
Q2_K 2.63 10.6 GB 13.4 GB Fits, but tight 11K 50.3
IQ2_XXS 2.06 8.5 GB 11.3 GB Runs comfortably 20K 61.0
IQ1_M 1.75 7.4 GB 10.2 GB Runs comfortably 24K 69.0

Also worth checking

DeepSeek-R1-Distill-Qwen 32B on every GPU
Full hardware tableEverything that runs on a RTX 5070 Ti
Full model tableWhy the KV cache is the part that bites
Guide