Rent a GPU

Can the NVIDIA RTX 5070 run Gemma 4 31B?

RTX 5070 × Gemma 4 31B8K context
NoDoes not fit in memory
Memory at Q4_K_M20.8 GB of 12 GB
Weights 19 GBKV cache 1.2 GBRuntime 0.6 GB
On the GPU54%of the Q4_K_M weights
In system RAM8.8 GBthe rest spills over
Offload speed~3.7tokens per second
  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, Gemma 4 31B needs about 20.8 GB, but the NVIDIA RTX 5070 has 12 GB usable. You could offload some layers to system RAM, but it will be very slow.

Try other context lengths

Run it on a rented GPU

On a cloud GPU with enough memory, the same command you'd use locally works:

ollama run gemma4:31b

At Q4_K_M it needs about 20.8 GB, so pick a GPU with at least 24 GB.

New to this? Step-by-step setup for the RTX 5070

Run it in the cloud

Too big for the NVIDIA RTX 5070? Run Gemma 4 31B 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 12 GB the NVIDIA RTX 5070 can use.

QuantQualityMemoryFits?Speed
FP16Full quality64.4 GBToo big—
Q8_0Near-lossless35 GBToo big—
Q6_KExcellent27.4 GBToo big—
Q5_K_MVery good24 GBToo big—
Q4_K_MGood, most popular20.8 GBToo big—
Q3_K_MNoticeably degraded17.1 GBToo big—
Q2_KHeavily degraded13.5 GBToo big—

Measured speed from real RTX 5070 owners

Tokens per second that people measured on their own NVIDIA RTX 5070, next to our estimate.

No one has measured Gemma 4 31B on the NVIDIA RTX 5070 yet. The speeds above are estimates. Run the benchmark and yours becomes the first real number on this page.

Have an RTX 5070? Add your real speed

One command times Gemma 4 31B 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 -s -- gemma4:31b

Windows (PowerShell)

$env:GPUFIT_MODEL="gemma4:31b"; irm https://gpufit.dev/bench.ps1 | iex

How the benchmark works

GPUs that run Gemma 4 31B comfortably

Check RTX 3090 prices

Models that run well on the NVIDIA RTX 5070

Frequently asked questions

Can NVIDIA RTX 5070 run Gemma 4 31B?

No. Even at Q4_K_M, Gemma 4 31B needs about 20.8 GB, but the NVIDIA RTX 5070 has 12 GB usable. You could offload some layers to system RAM, but it will be very slow.

How much VRAM does Gemma 4 31B need?

About 20.8 GB at Q4_K_M with 8,192 tokens of context, and about 64.4 GB at full FP16 precision.

How fast is Gemma 4 31B on the NVIDIA RTX 5070?

It does not fit in the NVIDIA RTX 5070's memory, so most layers would run from system RAM and generation would be very slow.