Yes — Devstral Small 24B (24B) runs at 35 tok/s on M4 Max with 48 GB RAM using Q4_K_M quantization via MLX. First token latency is 0.8s. A capable open-source LLM with 24B parameters.
LLMCheck measured Devstral Small 24B on M4 Max using the standard methodology: Q4_K_M quantization, 256-token input, 512-token output, 3 runs averaged on a freshly-booted system.
| Metric | Value |
|---|---|
| Tokens per second | 35 tok/s |
| Time to first token | 0.8s |
| Quantization | Q4_K_M |
| Minimum RAM | 48 GB |
| Recommended engine | MLX |
| Parameters | 24B |
| Benchmark date | 2026-05 |
Q4_K_M 24B MLX M4 Max
The recommended engine for Devstral Small 24B on M4 Max is MLX. Install with pip and pull the model:
Alternatively, you can use Ollama for a simpler setup:
| Chip | Speed | First Token | Min RAM | Engine |
|---|---|---|---|---|
| M5 Max | 45 tok/s | 0.6s | 128 GB | MLX |
| M5 Max | 38 tok/s | 0.7s | 64 GB | Ollama |
| M4 Pro | 26 tok/s | 1.0s | 32 GB | Ollama |
To run Devstral Small 24B on M4 Max you need:
See how Devstral Small 24B stacks up against other models on your specific Mac hardware.
Open Compare Tool Full Leaderboard