#1 · Tight but possible
Devstral (24B)
Best open model for coding agents. Designed for agentic coding workflows. 391K downloads.
codepower
Apple Silicon hardware guide
Best local LLMs for MacBook Pro M4 Pro 24GB
MacBook Pro M4 Pro 24GB with 24GB unified memory is a coding and reasoning models on the go machine. This page lists local AI models that fit its memory budget, with realistic performance expectations for LM Studio and similar runtimes.
Chip
M4 Pro
Unified memory
24GB
Compatible models
126
Best pick
Devstral (24B)
Quick answer
For MacBook Pro M4 Pro 24GB, start with Devstral (24B). Models marked “Comfortable” leave useful memory headroom; “Tight but possible” can work, but you should close other apps and prefer lower quantization.
Top compatible local LLMs
#2 · Good
Qwen 3 (14B)
The sweet spot. Incredible reasoning, coding and chat quality. The best model you can run on 16GB.
chatcodereasoningpowergeneral
#3 · Tight but possible
Gemma 4 26B A4B
Gemma 4 MoE flagship-for-workstations: 26B total with ~4B active parameters. 256K context and excellent quality-per-watt for local inference. Apache 2.0.
chatcodereasoningpowermultimodal
#4 · Tight but possible
Qwen 3 Coder (30B)
Qwen flagship coding model. Designed for agentic coding with 256K context. Outperforms Claude 3.5 Sonnet on SWE-bench. Apache 2.0.
codepowerquality
#5 · Good
Bonsai 27B
PrismML low-bit model derived from Qwen 3.6 27B. Official Apache 2.0 ternary (7.2GB deployed) and 1-bit (3.9GB) builds retain multimodal, reasoning and agentic capabilities through custom GGUF and MLX runtimes.
chatcodereasoningvisionagentic
#6 · Comfortable
LFM2.5-8B-A1B
Liquid AI hybrid model built for on-device assistants. 8.3B total / 1.5B active, 128K context, tool use, GGUF, ONNX, MLX, llama.cpp and LM Studio support. Open-weight under LFM 1.0.
chatcodereasoningspeedstandard
#7 · Good
GLM 4.5 Air (MoE)
Zhipu AI's efficient MoE powerhouse. 106B total parameters, only 14B active at inference — dense-model speed with much larger model quality. Clearly the best in the 16–24GB RAM range. Outperforms Llama 3.3 70B. Apache 2.0.
chatcodepowerqualitygeneral
#8 · Tight but possible
ZAYA1-8B
Zyphra's Apache-2.0 reasoning MoE: 8.4B total parameters with only ~760M active, 16 experts, 131K context, Compressed Convolutional Attention and strong math/code benchmarks. Experimental for local use today: currently needs Zyphra vLLM/Transformers forks; LM Studio/GGUF/MLX support is not yet verified.
chatcodereasoningmathexperimental
#9 · Comfortable
Granite 4.1 (8B)
IBM Granite 4.1 long-context instruct model. Apache 2.0, 131K context, tool calling, RAG, code tasks, multilingual dialog and business assistant workflows on normal 8-16 GB machines.
chatcodereasoningstandardgeneral
#10 · Comfortable
GLM 4.6 Air (12B)
Zhipu AI lightweight flagship. Strong bilingual CN/EN with hybrid thinking mode, 200K context and tool calling. Apache 2.0 — excellent alternative to Qwen 3.5 9B on modest GPUs.
chatcodereasoningstandardgeneral
#11 · Good
Apriel Nemotron 15B Thinker
ServiceNow x NVIDIA mid-size reasoner. Half the memory of 32B reasoners with comparable performance on MBPP, BFCL, GPQA. Strong enterprise fit. MIT licensed.
reasoningcodepowergeneral
#12 · Good
Ministral 3 14B Instruct
Mistral AI larger Ministral 3 instruct model. Apache 2.0, official GGUF availability, better quality ceiling than the 3B/8B variants while staying practical on 16-32GB workstations.
chatvisionpowerreasoningmultilingual
Buying note
This page is about local AI fit, not a live price tracker. Prices and availability change. If an Amazon link is present, it may be an affiliate link that supports LocalClaw at no extra cost.