Run Devstral Small 24B on M4 Pro

Yes — Devstral Small 24B (24B) runs at 26 tok/s on M4 Pro with 32 GB RAM using Q4_K_M quantization via Ollama. First token latency is 1.0s. A capable open-source LLM with 24B parameters.

Speed
26
tok/s
First Token
1.0
seconds
RAM Needed
32
GB minimum
Engine
Ollama
recommended

Benchmark Details

LLMCheck measured Devstral Small 24B on M4 Pro using the standard methodology: Q4_K_M quantization, 256-token input, 512-token output, 3 runs averaged on a freshly-booted system.

MetricValue
Tokens per second26 tok/s
Time to first token1.0s
QuantizationQ4_K_M
Minimum RAM32 GB
Recommended engineOllama
Parameters24B
Benchmark date2026-05

Q4_K_M 24B Ollama M4 Pro

Setup Guide: Run Devstral Small 24B on M4 Pro

The recommended engine for Devstral Small 24B on M4 Pro is Ollama. Install Ollama, then pull the model:

ollama run devstral-small-24b

Ollama handles quantization automatically — it will download the Q4_K_M variant (~32 GB) and start an interactive chat session.

Performance on Other Apple Silicon Chips

ChipSpeedFirst TokenMin RAMEngine
M5 Max 45 tok/s 0.6s 128 GB MLX
M5 Max 38 tok/s 0.7s 64 GB Ollama
M4 Max 35 tok/s 0.8s 48 GB MLX

System Requirements

To run Devstral Small 24B on M4 Pro you need:

Compare More Models

See how Devstral Small 24B stacks up against other models on your specific Mac hardware.

Open Compare Tool Full Leaderboard