With 32 GB unified memory (about 24.0 GB usable by the GPU), these are the best open models that fit. The top pick is Qwen3 30B-A3B (MoE) at Q4_K_M (~20.4 GB, ~23 tok/s): ollama run qwen3:30b.
Best for coding: Qwen3-Coder 30B-A3B (MoE) (qwen3-coder:30b), Qwen3 30B-A3B (MoE) (qwen3:30b), gpt-oss 20B (MoE, MXFP4) (gpt-oss:20b).
Runtimes available for this hardware: Ollama, LM Studio, llama.cpp, MLX.
Open the interactive checker for M4 32GB| # | Model | Quant | Memory needed | Fit | Speed | Command |
|---|---|---|---|---|---|---|
| 1 | Qwen3 30B-A3B (MoE) | Q4_K_M | 20.4 GB | fits in fast memory | ~23 tok/s | ollama run qwen3:30b |
| 2 | Qwen3-Coder 30B-A3B (MoE) | Q4_K_M | 20.4 GB | fits in fast memory | ~23 tok/s | ollama run qwen3-coder:30b |
| 3 | gpt-oss 20B (MoE, MXFP4) | MXFP4 | 15.1 GB | fits in fast memory | ~20 tok/s | ollama run gpt-oss:20b |
| 4 | Mistral Small 24B | Q4_K_M | 15.8 GB | fits in fast memory | ~7 tok/s | ollama run mistral-small:24b |
| 5 | Gemma 3 27B | Q4_K_M | 19 GB | fits in fast memory | ~6 tok/s | ollama run gemma3:27b |
| 6 | Qwen3 32B | Q4_K_M | 22.2 GB | fits in fast memory | ~5 tok/s | ollama run qwen3:32b |
| 7 | DeepSeek R1 Distill Qwen 32B | Q4_K_M | 22.2 GB | fits in fast memory | ~5 tok/s | ollama run deepseek-r1:32b |
| 8 | Nimble v3 9B (Bespoke Labs, decision) | Q8_0 | 10.6 GB | fits in fast memory | ~10 tok/s | ollama run nimble-v3:9b-q8_0 |
| 9 | Qwen3 8B | Q8_0 | 9.9 GB | fits in fast memory | ~11 tok/s | ollama run qwen3:8b-q8_0 |
| 10 | DeepSeek R1 0528 Qwen3 8B | Q8_0 | 9.9 GB | fits in fast memory | ~11 tok/s | ollama run deepseek-r1:8b-0528-qwen3-q8_0 |
| 11 | Clef 27B (Cloudflare, decision) | Default | 19.9 GB | fits in fast memory | ~5 tok/s | ollama run clef:27b |
| 12 | Qwen3.8 27B | Default | 19.9 GB | fits in fast memory | ~5 tok/s | ollama run qwen3.8:27b |
Estimates, not benchmarks. Model sizes are ollama.com download sizes; memory needed adds headroom for ~8k tokens of context; speeds are rough figures from memory bandwidth. Data updated 2026-10-14.
A project by OnDevice Labs · llms.txt