Only with CPU offload

Devstral Small 2 24B at Q4_K_M needs 15.3 GB but only 14.4 GB is addressable, so about 7% of the layers would stream from system RAM at roughly 60 GB/s. Expect around 20 tokens per second — usable for batch work, painful for chat.

93% on GPU 8K context Q4_K_M · 13.5 GB Apache 2.0 Released Dec 2025 Vision Not in the Ollama library

Built for software-engineering agents (OpenHands, Cline). Dense 24B, 384K window. Not in the Ollama library.

Hardware requirements What hardware do I need to run Devstral Small 2 24B? →

At Q4_K_M with an 8K context: a 20 GB GPU (RTX 4000 Ada) or a 32 GB Mac (M4). Minimum, recommended, high-quality and long-context tiers, with all 114 devices checked at every quantisation.

The VRAM budget

weights 13.5 GB
Weights 13.5 GB KV cache @ 8K 1.25 GB Runtime overhead 0.6 GB Over budget 0.9 GB past 14.4 GB

Quantisation ladder

QuantWeightsTotal @ 8KMax contextTok/sQualityFit
Q8_0 23.7 GB 25.6 GB — ~3.0 −0.1% ppl 11.2 GB over
Q6_K 18.3 GB 20.2 GB — ~5.5 −0.4% ppl 5.8 GB over
Q5_K_M 15.8 GB 17.7 GB — ~8.8 −0.8% ppl 3.3 GB over
Q4_K_M 13.5 GB 15.3 GB 1K ~20 −1.9% ppl 0.9 GB over
Q3_K_M 10.9 GB 12.8 GB 18K 50 −5.4% ppl Long context
Q2_K 9.4 GB 11.2 GB 28K 58 −15% ppl Long context

Quality is the published perplexity delta against f16 weights. Max context assumes an f16 KV cache; q8_0 roughly doubles it.

How to run it

terminal
$ llama-server \
    -hf mistralai/Devstral-Small-2-24B-Instruct-2512:Q4_K_M \
    -c 8192 -ngl 37

The engine underneath most of the others. Every knob is exposed. More on llama.cpp.

01Download is 13.5 GB. Keep it on an SSD — a first load off a spinning disk takes minutes.
02Close anything else holding VRAM. A browser with hardware acceleration can sit on 1–2 GB.
03If it falls back to CPU silently, drop the context first, then step down a quantisation.
See all models for this rig Compare with another model

Related models on a GeForce RTX 5070 Ti

ModelTotalTok/s est.
Mistral Small 3.2 24B
23.6B · similar size
Q4_K_M · 15.1 GB ~23 CPU offload
Gemma 4 26B-A4B
26.5B MoE · similar size
Q4_K_M · 16.0 GB ~39 CPU offload
Gemma 3 27B
27.4B · similar size
Q4_K_M · 17.0 GB ~10 CPU offload
gpt-oss 20B
20.9B MoE · similar size
MXFP4 · 11.6 GB 112 Runs great
Qwen3.8 27B
27.8B · similar size
Q4_K_M · 16.7 GB ~11 CPU offload
Qwen3.6 27B
27.8B · similar size
Q4_K_M · 16.7 GB ~11 CPU offload
Granite 4.1 30B
28.9B · similar size
Q4_K_M · 18.9 GB ~6.9 CPU offload

Every model against the GeForce RTX 5070 Ti →

Devstral Small 2 24B on other hardware

DeviceTotalTok/s est.
GeForce RTX 4060 Laptop
7.0 GB usable · 256 GB/s
15.3 GB ~3.8 CPU offload
GeForce RTX 3060 12 GB
10.6 GB usable · 360 GB/s
15.3 GB ~5.9 CPU offload
M4 · 16 GB
10.7 GB usable · 120 GB/s
15.3 GB — Won't fit
RTX 4000 Ada
18.4 GB usable · 360 GB/s
15.3 GB 16 Runs great
GeForce RTX 4090
22.4 GB usable · 1008 GB/s
15.3 GB 45 Runs great
Radeon RX 7900 XTX
22.4 GB usable · 960 GB/s
15.3 GB 43 Runs great
M4 · 32 GB
24.0 GB usable · 120 GB/s
15.3 GB 5.0 Runs great
M3 Max · 64 GB
48.0 GB usable · 400 GB/s
15.3 GB 17 Runs great

All 114 devices, by tier →