LocalForge
Back to guide
MODEL / MISTRAL AI

Devstral Small 2 24B

The 256K maximum context is not a practical default on consumer memory.

Primary source
Total parameters24B
Estimated memory~18.5 GBGGUF Q4 / BF16
Context256KText + Image
LicenseApache-2.0Multilingual
01 / BEST FOR
  • coding agents
  • repository work
  • tool use
02 / ESTIMATED FIT
Apple silicon · 32 GBTightNVIDIA · 24 GBTightApple silicon · 64 GBComfortableWorkstation · 48 GBComfortableDatacenter · 80 GBComfortable
03 / DEPLOY WITH
OllamaEasyollama run <model-tag>
llama.cppMediumllama-server -m /path/to/verified-model.gguf --port 8080
LM StudioEasySearch “mistralai/Devstral-Small-2-24B-Instruct-2512” → Load a Q4/INT4 build → Start Server
MLX LMMediummlx_lm.chat --model mistralai/Devstral-Small-2-24B-Instruct-2512
Watch out

The 256K maximum context is not a practical default on consumer memory.