Can I Run Codestral on 48 GB system RAM?
Superseded model. Codestral has been superseded by Devstral. This page is kept for reference; the newer family is a better starting point.
View Devstral →
Written by Jakub Rusinowski · Last updated May 29, 2024
Yes — comfortably
Yes, comfortably — Codestral 22B at Q3_K_M needs about 12.1 GB of the 38.4 GB usable on 48 GB system RAM, leaving ~26.3 GB spare and running at ~6.4 tok/s (estimated), with room for about 32,768 tokens of context.
Confidence: high · Recommended quantization: Q3_K_M · Estimated speed: ~6.4 tok/s
48 GB system RAM — what it gives a model
| Usable memory for models | 38.4 GB |
| Memory bandwidth | 90 GB/s |
Codestral on 48 GB system RAM: memory by quantization
| Quant | Memory needed | Fits 38.4 GB? | Max context | Est. speed | Download |
|---|
| F16 | 47.1 GB | ✗ No | — | — | 44.4 GB |
| Q8_0 | 26.3 GB | ✓ Yes | 32K | ~2.7 tok/s | 23.6 GB |
| Q6_K | 20.9 GB | ✓ Yes | 32K | ~3.5 tok/s | 18.2 GB |
| Q5_K_M | 18.4 GB | ✓ Yes | 32K | ~4 tok/s | 15.7 GB |
| Q4_K_M | 16.1 GB | ✓ Yes | 32K | ~4.6 tok/s | 13.4 GB |
| Q3_K_M | 12.1 GB | ✓ Yes | 32K | ~6.4 tok/s | 9.5 GB |
| Q2_K | 10 GB | ✓ Yes | 32K | ~8 tok/s | 7.3 GB |
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.
- These figures assume CPU-only inference. Any discrete GPU, even an 8 GB one, will be several times faster for models that fit in its VRAM.
Recommended setup
llama.cpp (CPU build) or Ollama — both run without a GPU
How these numbers are calculated
- Memory need = quantized weights + KV cache at 8,192 tokens (f16) + 0.8 GB runtime overhead.
- 38.4 GB of the 48 GB is treated as usable for model weights (80% — the rest is the OS and running applications).
- DDR5-5600 dual channel at 89.6 GB/s peak. CPU decode is assumed to sustain 35% of that peak, because CPU inference is not purely bandwidth-bound — it also spends real time in compute and thread synchronisation. This figure is an assumption, not a fitted constant: no CPU measurement is in the calibration set.
- CPU-only inference: no GPU is assumed. A GPU of any size will beat these figures substantially.
- 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
Can I run Codestral on 48 GB system RAM?
Yes, comfortably — Codestral 22B at Q3_K_M needs about 12.1 GB of the 38.4 GB usable on 48 GB system RAM, leaving ~26.3 GB spare and running at ~6.4 tok/s (estimated), with room for about 32,768 tokens of context.
Which quantization of Codestral should I use on 48 GB system RAM?
Q3_K_M — it needs about 12.1 GB of the 38.4 GB available, downloads as roughly 9.5 GB, and runs at an estimated 6.4 tokens/sec with up to 32K of context.
What limits Codestral on 48 GB system RAM?
Nothing binding — the model fits with headroom and generates at a usable speed on this hardware.
Which runtime should I use?
llama.cpp (CPU build) or Ollama — both run without a GPU
Other RAM Capacities
Other Models on 48 GB system RAM
Codestral on GPUs
What This Model Is Good At
Model & Tools
← Can I Run It? | Codestral model page | Check your hardware