Can I Run DeepSeek R1 0528 Qwen3 8B on a NVIDIA RTX 5090?
Runs at Q8_0 — near-lossless quality.
12 quantizations fit your 32.0GB
| Quant | Min VRAM | Recommended | File size | Headroom |
|---|---|---|---|---|
| Q8_0BEST | 9.7 GB | 11.2 GB | 8.7 GB | +22.3 GB |
| Q6_K | 7.8 GB | 9.3 GB | 6.7 GB | +24.3 GB |
| Q5_K_M | 6.8 GB | 8.3 GB | 5.8 GB | +25.2 GB |
| Q5_K_S | 6.7 GB | 8.2 GB | 5.7 GB | +25.3 GB |
| Q4_1 | 6.1 GB | 7.6 GB | 5.3 GB | +25.9 GB |
| Q4_K_M | 6.0 GB | 7.5 GB | 5.0 GB | +26.0 GB |
| Q4_K_S | 5.7 GB | 7.2 GB | 4.8 GB | +26.3 GB |
| Q4_0 | 5.6 GB | 7.1 GB | 4.8 GB | +26.4 GB |
| Q3_K_L | 4.7 GB | 6.2 GB | 4.4 GB | +27.4 GB |
| Q3_K_M | 4.4 GB | 5.9 GB | 4.1 GB | +27.6 GB |
| Q3_K_S | 4.2 GB | 5.7 GB | 3.8 GB | +27.8 GB |
| Q2_K | 3.7 GB | 5.2 GB | 3.3 GB | +28.3 GB |
Try it in the cloud first
Don't want to download DeepSeek R1 0528 Qwen3 8B just to try it? Use a hosted API or rent a GPU by the second.
Affiliate links — we earn a commission at no cost to you.
All quant variants, benchmark scores, and use-case tags.
Top-ranked open-source models that fit in 32.0GB.
FAQ
Can the NVIDIA RTX 5090 run DeepSeek R1 0528 Qwen3 8B?
Yes. The NVIDIA RTX 5090's 32.0GB of VRAM is enough to run DeepSeek R1 0528 Qwen3 8B at Q8_0 quantization (9.7GB required).
What's the best quantization to use?
Q8_0 is the highest-precision quantization that fits in your 32.0GB. It uses about 9.7GB of memory and 11.2GB recommended for comfortable inference.
What if I need more headroom for context length?
KV cache memory grows with context length. The numbers above assume a baseline 2K-4K context. For long-context use (32K+), add another 2-6GB depending on the model architecture.