Can I Run Qwen3 32B on an RTX 5090?

Yes. At Q4_K_M and an 8K context it needs about 23.2 GB, which is 73 percent of the RTX 5090’s 32 GB.

Check your own setup with the Can My GPU Run This AI Model? calculator. The numbers below come from the same data.

Memory needed on the RTX 5090

Quantization Weights Total at 8K (est.) On 32 GB Speed (est., tok/s)
Q4_K_M 20.1 GB 23.2 GB Fits 54 to 71
Q5_K_M 23.3 GB 26.5 GB Fits 46 to 61
Q6_K 27.0 GB 30.2 GB Tight 40 to 53
Q8_0 34.8 GB 38.0 GB No –

The highest quantization that still fits comfortably is Q5_K_M.

Context length matters

Context KV cache Total at Q4_K_M (est.) On 32 GB
4K 1.07 GB 22.1 GB Fits
8K 2.15 GB 23.2 GB Fits
16K 4.29 GB 25.4 GB Fits
32K 8.59 GB 29.7 GB Tight

What speed to expect

With 1792 GB/s of bandwidth, expect about 54 to 71 tokens per second at Q4_K_M. We did not find a reliable measured result for this exact card and model, so treat speeds as estimates that could be off by 10 to 20 percent.

Estimates: weights = parameters × bits per weight; KV cache from the model’s published architecture; about 1 GB of runtime overhead. “Fits” means at most 92 percent of VRAM, “Tight” is 92 to 100 percent. Speed is memory bandwidth divided by model size at 60 to 80 percent efficiency. These are planning estimates, not our own benchmarks.

Verdict

Run it at Q4_K_M and enjoy. For a wider view, see what GPU you need for Qwen3 32B and everything the RTX 5090 can run.

Frequently asked questions

Can the RTX 5090 run Qwen3 32B at Q8?

Q8_0 needs about 38.0 GB at 8K context: no on 32 GB.

How much context can the RTX 5090 handle with Qwen3 32B?

At Q4_K_M, 32K context needs about 29.7 GB in total (tight on 32 GB).

Sources and method

  • Model architecture: Qwen3 32B config.json on Hugging Face.
  • GPU specs: PNY NVIDIA GeForce linecard, 7 Jan 2026 (pny.com consumer-geforce-linecard.pdf).
  • All VRAM and speed figures are estimates, not our own benchmarks.

Last updated 10 Nov 2026.

Spynium


Leave a Comment