Rent a GPU

Can the NVIDIA RTX 4080 run DeepSeek V4 Pro 1.6T-A49B?

RTX 4080 × DeepSeek V4 Pro 1.6T-A49B8K context
NoDoes not fit in memory
Memory at Q4_K_M971.2 GB of 16 GB
Weights 970 GBKV cache 0.6 GBRuntime 0.6 GB
  1. FP16
  2. Q8_0
  3. Q6_K
  4. Q5_K_M
  5. Q4_K_M
  6. Q3_K_M
  7. Q2_K

No. Even at Q4_K_M, DeepSeek V4 Pro 1.6T-A49B needs about 971.2 GB, but the NVIDIA RTX 4080 has 16 GB usable. You could offload some layers to system RAM, but it will be very slow.

Try other context lengths

Run it in the cloud

Too big for the NVIDIA RTX 4080? Run DeepSeek V4 Pro 1.6T-A49B on a rented GPU instead.

Rent a cloud GPU by the hour. No upfront cost, stop any time. Not sure whether to rent or buy? Compare the cost.

Rent a GPU on RunPodRent a GPU on Vast.ai

Memory needed by quantization

At 8,192 tokens of context. The black line marks the 16 GB the NVIDIA RTX 4080 can use.

QuantQualityMemoryFits?Speed
FP16Full quality3201.2 GBToo big—
Q8_0Near-lossless1701.2 GBToo big—
Q6_KExcellent1313.2 GBToo big—
Q5_K_MVery good1139.2 GBToo big—
Q4_K_MGood, most popular971.2 GBToo big—
Q3_K_MNoticeably degraded783.2 GBToo big—
Q2_KHeavily degraded601.2 GBToo big—

Measured speed from real RTX 4080 owners

Tokens per second that people measured on their own NVIDIA RTX 4080, next to our estimate.

No one has measured DeepSeek V4 Pro 1.6T-A49B on the NVIDIA RTX 4080 yet. The speeds above are estimates. Run the benchmark and yours becomes the first real number on this page.

Have an RTX 4080? Add your real speed

One command times DeepSeek V4 Pro 1.6T-A49B in your own Ollama, shows you exactly what it would send, and asks before sending anything. No sign-up, no personal data.

macOS and Linux

curl -fsSL https://gpufit.dev/bench.sh | sh

Windows (PowerShell)

irm https://gpufit.dev/bench.ps1 | iex

How the benchmark works

Models that run well on the NVIDIA RTX 4080

Frequently asked questions

Can NVIDIA RTX 4080 run DeepSeek V4 Pro 1.6T-A49B?

No. Even at Q4_K_M, DeepSeek V4 Pro 1.6T-A49B needs about 971.2 GB, but the NVIDIA RTX 4080 has 16 GB usable. You could offload some layers to system RAM, but it will be very slow.

How much VRAM does DeepSeek V4 Pro 1.6T-A49B need?

About 971.2 GB at Q4_K_M with 8,192 tokens of context, and about 3201.2 GB at full FP16 precision.

How fast is DeepSeek V4 Pro 1.6T-A49B on the NVIDIA RTX 4080?

It does not fit in the NVIDIA RTX 4080's memory, so most layers would run from system RAM and generation would be very slow.