Can I Run Codestral on RTX 4090 Laptop (16 GB VRAM, 32 GB RAM)?

Autor: Jakub Rusinowski · Ostatnia aktualizacja: 29 maja 2024

Yes

Yes — Codestral 22B at Q3_K_M needs about 12.1 GB of the 16 GB usable on RTX 4090 Laptop (16 GB VRAM, 32 GB RAM) (~3.9 GB spare), at ~38 tok/s (estimated), with room for about 16,384 tokens of context.

Confidence: high · Recommended quantization: Q3_K_M · Estimated speed: ~38 tok/s

See what else this hardware can run →

Uruchom w chmurze RTX 4090 w RunPod — od $0.34/godz. · stawka sprawdzona 2026-07

lub porównaj w Vast.ai od $0.35/godz. (typical low · varies)

Jako partner Amazon zarabiamy na kwalifikujących się zakupach. Linki do chmurowych GPU są linkami polecającymi — możemy otrzymać prowizję bez dodatkowych kosztów dla Ciebie.

RTX 4090 Laptop (16 GB VRAM, 32 GB RAM) — what it gives a model

Usable memory for models	16 GB
Memory bandwidth	576 GB/s
Form factor	Laptop
Operating system	Windows or Linux
Memory upgradeable	Yes

Codestral on RTX 4090 Laptop (16 GB VRAM, 32 GB RAM): memory by quantization

Quant	Memory needed	Fits 16 GB?	Max context	Est. speed	Download
F16	47.1 GB	✗ No	—	—	44.4 GB
Q8_0	26.3 GB	✗ No	—	—	23.6 GB
Q6_K	20.9 GB	✗ No	—	—	18.2 GB
Q5_K_M	18.4 GB	✗ No	—	—	15.7 GB
Q4_K_M	16.1 GB	✗ No	—	—	13.4 GB
Q3_K_M	12.1 GB	✓ Yes	16K	~38 tok/s	9.5 GB
Q2_K	10 GB	✓ Yes	32K	~46.9 tok/s	7.3 GB

Superseded model. Codestral has been superseded by Devstral. This page is kept for reference; the newer family is a better starting point. Devstral needs at least 16.6 GB at Q4_K_M with an 8K context, so it does not fit the RTX 4090 Laptop (16 GB VRAM, 32 GB RAM).

What to watch out for

Q3_K_M is the only quantization that fits, and it is a heavily degraded one — expect noticeably worse output than the same model at Q4_K_M. A smaller model at Q4 is usually the better trade.
1 larger variant of Codestral does not fit and would need CPU offload or different hardware.

RTX 4090 laptop limitations

A mobile RTX 4090 carries 16 GB, not the desktop card's 24 GB, and is closer to a desktop 4080 in throughput — which is why it is modelled against that chip here.
Sustained throughput depends on the chassis power limit; thin laptops throttle well below the quoted figures.

Recommended setup

Ollama or llama.cpp (CUDA/ROCm) — vLLM if you need concurrent serving

How these numbers are calculated

Memory need = quantized weights + KV cache at 8,192 tokens (f16) + 0.8 GB runtime overhead.
16 GB of VRAM on the NVIDIA GeForce RTX 4090 Laptop GPU at 576 GB/s.
32 GB of system RAM available for CPU offload when a model exceeds VRAM.
Throughput is a memory-bandwidth roofline estimate, not a measurement. It is labelled "estimated" everywhere it appears.
KV cache is computed from this model's published attention configuration.

FAQ