Can I Run Qwen3.8 27B on NVIDIA GeForce RTX 5090?
Autor: Jakub Rusinowski · Ostatnia aktualizacja: 29 września 2026
Yes. Qwen3.8 27B needs 18.3 GB at 8K context (16.8 GB weights + 1.5 GB KV cache/overhead) at Q4_K_M and the NVIDIA GeForce RTX 5090 has 32 GB, leaving about 13.7 GB of headroom, at roughly 74 tok/s (estimated) with room for up to 128K of context.
See what else this hardware can run →
The numbers
| Model | Qwen3.8 27B |
| Parameters | 27B |
| Quantization | Q4_K_M |
| VRAM needed (8K context) | 18.3 GB |
| NVIDIA GeForce RTX 5090 VRAM | 32 GB |
| Verdict | Runs |
| Max context | 128K |
| Estimated speed | ~74 tok/s |
Qwen3.8 27B on the NVIDIA GeForce RTX 5090: VRAM by quantization
| Quant | VRAM needed | Fits 32 GB? | Max context | Whole PC |
|---|---|---|---|---|
| F16 | 57.1 GB | ✗ No | — | — |
| Q8_0 | 31 GB | ✓ Yes | 16K | Build for this |
| Q6_K | 24.3 GB | ✓ Yes | 64K | — |
| Q5_K_M | 21.2 GB | ✓ Yes | 128K | Build for this |
| Q4_K_M | 18.3 GB | ✓ Yes | 128K | Build for this |
| Q3_K_M | 13.3 GB | ✓ Yes | 256K | — |
| Q2_K | 10.6 GB | ✓ Yes | 256K | — |
Assumes an 8K-token context with an f16 KV cache. A longer window needs more; a quantized KV cache needs less. “Max context” is the largest window that still fits in 32 GB. Figures are estimates from parameter count, quantization and memory bandwidth — the analyzer lets you tune KV-cache quant and context.
Context costs VRAM too. At Q4_K_M on the NVIDIA GeForce RTX 5090: 8K 18.3 GB · 32K 19.9 GB · 128K 26.3 GB. Every window this model publishes still fits 32 GB.
What to do next
Nothing to change — run it at Q4_K_M. If you want more capability, check whether the next size up also fits.
No compatible GPU? Qwen3.8 27B on 16 GB of system RAM, CPU only: runs.
Running Qwen3.8 27B on this card:
At 2 hrs/day, buying (~$1,999) beats renting at $0.34/hr after about 8.2 years.
Affiliate links — we may earn a commission if you sign up, at no extra cost to you.
Cloud rates verified 2026-07 — estimates, and marketplace prices vary. Buying price is GPU MSRP only, not a full PC.
Other Qwen3.8 sizes on the NVIDIA GeForce RTX 5090
- Qwen3.8-Max — does not fit (1457.5 GB)
- Qwen3.8-Flash-Next — does not fit (109.7 GB)
Qwen3.8 27B on nearby hardware
- Apple M2 (24 GB) — runs at a lower quant
- Apple M3 Pro (36 GB) — runs
All Qwen3.8 sizes on the NVIDIA GeForce RTX 5090 | Qwen3.8 27B model page | NVIDIA GeForce RTX 5090 GPU page | VRAM calculator | Check your hardware