Devstral Small 2 (24B) on Marvell OCTEON TX2 CN9130 NAS (32 GB)

Can the Marvell OCTEON TX2 CN9130 NAS (32 GB) run Devstral Small 2 (24B)?

Barely: it runs very slowly

Very slow. At Q5_K_M it needs about 21.1 GB of 26.2 GB usable, with an estimated speed of 0.3 tok/s.

Best quant: Q5_K_M
Generation speed: 0.3 tok/s
Prompt processing: 0.5 tok/s
First token (1K prompt): 1,966.9 s
Largest safe context: 38K

Best app to run it on the Marvell OCTEON TX2 CN9130 NAS (32 GB)

On Linux on ARM boards

Ollama

Command line + local API · runs on llama.cpp (ggml), MLX (Apple silicon, preview)

One command with Ollama’s own build of this model.
Serves an OpenAI-compatible API that other apps and coding tools can use.

ollama run devstral-small-2:24b

llama.cpp Usually the first to support new model architectures.

How we choose: the files published for this model, which apps run in a container on a NAS without a graphics card (models run on its processor), and whether the model fits fully in the memory models can use. Details.

Download Devstral Small 2 (24B)

Recommended for the Marvell OCTEON TX2 CN9130 NAS (32 GB): Q5_K_M. Every other quantization has its own download link in the table below.

Quickest · Ollama

One command downloads a ready-to-run version and starts a chat.

Get Ollama at ollama.com/download (opens in a new tab) · model page (opens in a new tab)

ollama run devstral-small-2:24b

Quantized files · GGUF

Works in LM Studio, llama.cpp, Jan and Ollama. No sign-in needed.

Download the Q5_K_M file (opens in a new tab)

huggingface.co/unsloth/Devstral-Small-2-24B-Instruct-2512-GGUF (opens in a new tab)

ollama run hf.co/unsloth/Devstral-Small-2-24B-Instruct-2512-GGUF:Q5_K_M

Original weights · Hugging Face

Full-precision files from Mistral AI, for fine-tuning, vLLM, Transformers.

huggingface.co/mistralai/Devstral-Small-2-24B-Instruct-2512 (opens in a new tab)

No sign-in needed.

Official repo · GitHub

Reference code, inference scripts and docs from Mistral AI. The weights themselves download from Hugging Face.

github.com/mistralai/mistral-inference (opens in a new tab)

Mistral AI’s official inference code.

git clone https://github.com/mistralai/mistral-inference

Links checked 3 days ago against Hugging Face, Ollama and GitHub.

How it’s rated

Quality 73Great quality (A) · 73 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking.: How capable the model itself is at Q5_K_M, from its editorial quality tier scaled by the quantization or precision. It’s the same on any hardware that runs this variant. Scale: A+ 85+ · A 70+ · B 55+ · C 40+ · D under 40.
Runs Very slow: How well it runs on the Marvell OCTEON TX2 CN9130 NAS (32 GB): speed, capped by how well it fits in memory. Runs great: 30+ tokens/s (or near the image, video or decision speed target) · Runs well: 10+ tokens/s · Slow: 3+ tokens/s · Very slow: under 3 · Won’t run: doesn’t fit even with offload. A tight fit or RAM offload caps it at Runs well; heavy offload caps it at Slow.

Match score

To order lists, each result also gets a match score out of 100 from four parts. It favors the most capable model that still runs well.

Model qualityEditorial quality tier of the model, scaled by the quantization or precision used.	33.2 / 45
Memory fitFull marks when the model fits in the share of the NAS’s memory that models can use (the rest stays with the NAS system and its storage services); fewer when it goes past that limit.	25.0 / 25
SpeedGeneration speed against a 40 tokens/s target for LLMs, or render time against a target for image and video.	1.3 / 15
Memory useRewards using the memory you have, so a bigger, better variant beats a tiny one that leaves most memory idle.	15.0 / 15
Match score	74.4 / 100

The match score orders results; it is not shown as a grade. Quality and Runs are the two ratings in the tables.

Every quantization on the Marvell OCTEON TX2 CN9130 NAS (32 GB)

Quant	Memory needed	Available	Speed	Runs	Quality	Download
`Q2_K`	13.6 GB	26.2 GB	0.6 tok/s	Very slow	53Fair quality (C) · 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`Q3_K_M`	15.8 GB	26.2 GB	0.4 tok/s	Very slow	63Good quality (B) · 63 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`Q4_K_M`	18.7 GB	26.2 GB	0.3 tok/s	Very slow	72Great quality (A) · 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`Q5_K_M`	21.1 GB	26.2 GB	0.3 tok/s	Very slow	73Great quality (A) · 73 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`Q6_K`	23.7 GB	26.2 GB	0.2 tok/s	Very slowtight fit	75Great quality (A) · 75 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`Q8_0`	29.5 GB	26.2 GB	—	Won't run	76Great quality (A) · 76 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`F16`	52.1 GB	26.2 GB	—	Won't run	76Great quality (A) · 76 of 100: how capable the model is at F16, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab) Original weights (opens in a new tab)

Memory figures assume 8K of context. Longer conversations or documents grow the KV cache.

Other hardware that runs it

75Great quality (A) · 75 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 5090 Q6_K · 63.6 tok/s · Runs great
73Great quality (A) · 73 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 4090 Q5_K_M · 42.6 tok/s · Runs great
73Great quality (A) · 73 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 3090 Q5_K_M · 39.5 tok/s · Runs great
72Great quality (A) · 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. AMD Radeon RX 7900 XTX Q4_K_M · 42.3 tok/s · Runs great
63Good quality (B) · 63 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 4070 Ti SUPER Q3_K_M · 44.4 tok/s · Runs well (tight fit)
63Good quality (B) · 63 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 4080 SUPER Q3_K_M · 48.6 tok/s · Runs well (tight fit)

Quick answers

How much memory does Devstral Small 2 (24B) need on the Marvell OCTEON TX2 CN9130 NAS 32GB?

At Q5_K_M it needs about 21.1 GB. The Marvell OCTEON TX2 CN9130 NAS 32GB has 26.2 GB usable.

How fast is Devstral Small 2 (24B) on the Marvell OCTEON TX2 CN9130 NAS 32GB?

About 0.3 tok/s at Q5_K_M, an estimate from the hardware’s memory bandwidth and the size of the model.

Which quantization of Devstral Small 2 (24B) should I use on the Marvell OCTEON TX2 CN9130 NAS 32GB?

Q5_K_M. Of the versions that fit, it is the one that runs best; when two run equally well, the check prefers the higher-rated one.

Estimates come from published hardware specs and our open formulas, not from running the model on your machine. Real speed varies with runtime, drivers and settings. Numbers marked “Measured” come from a linked benchmark.