AI Index

Open-weight models that run on Apple M4 Pro · 24 GB

  • Unified memory 24 GB
  • Bandwidth 273 GB/s
  • Usable 16.0 GB

Specs: Memory · Mac mini (2024) - Tech Specs - Apple Support · Bandwidth · Mac mini (2024) - Tech Specs - Apple Support

FitModelParameters / activePrecisionMemoryEst. speedLongest contextRuns as
Runs wellGemma 4 26B A4B IT Google25.8B / 3.8BIQ4_XS13.9 / 16.0 GB53–80 tok/s61KUnified memory
Runs wellERNIE-4.5-21B-A3B-Thinking Baidu21.8B / 3BQ4_K_M13.4 / 16.0 GB60–89 tok/s55KUnified memory
Runs wellGPT OSS 20B OpenAI20.9B / 3.6BMXFP4 (native)maker’s file13.6 / 16.0 GB53–80 tok/s108KUnified memory
Runs wellQwen2.5-14B-Instruct-1M Alibaba Qwen14.8BQ4_K_M10.4 / 16.0 GB13–19 tok/s38KUnified memory
Runs wellQwen2.5-14B-Instruct Alibaba Qwen14.8BQ4_K_Mmaker’s file10.4 / 16.0 GB13–19 tok/s32KUnified memory
Runs wellDeepSeek-R1-Distill-Qwen-14B DeepSeek14.8BQ4_K_M10.4 / 16.0 GB13–19 tok/s32KUnified memory
Runs wellQwen3-14B Alibaba Qwen14.8BQ4_K_Mmaker’s file10.1 / 16.0 GB13–20 tok/s128KUnified memory
Runs wellGemma 3 12B IT Google12.2BQ4_K_M10.2 / 16.0 GB13–20 tok/s24KUnified memory
Runs wellQwen2.5-Omni-7B Alibaba Qwen10.7BQ6_K9.1 / 16.0 GB15–22 tok/s32KUnified memory
Runs wellGLM-4.6V-Flash Zhipu AI10.3BFP810.4 / 16.0 GB13–19 tok/s125KUnified memory
Runs wellQwen3.5-9B Alibaba Qwen9.7BQ8_010.3 / 16.0 GB13–19 tok/s190KUnified memory
Runs wellQianfan-VL-8B Baidu8.8BQ8_010.2 / 16.0 GB13–20 tok/s32KUnified memory
Runs wellGranite Switch 4.1 8B IBM8.8BFP810.0 / 16.0 GB13–20 tok/s44KUnified memory
Runs wellQwen3-VL-8B-Instruct Alibaba Qwen8.8BQ8_0maker’s file9.7 / 16.0 GB14–21 tok/s52KUnified memory
Runs wellQwen3-VL-8B-Thinking Alibaba Qwen8.8BQ8_0maker’s file9.7 / 16.0 GB14–21 tok/s52KUnified memory
Runs wellpplx-embed-v2-context-9b Perplexity8.4BQ8_09.1 / 16.0 GB15–22 tok/s230KUnified memory
Runs wellQwen2.5-VL-7B-Instruct Alibaba Qwen8.3BQ8_09.1 / 16.0 GB15–22 tok/s125KUnified memory
Runs wellQwen3-8B Alibaba Qwen8.2BMLX 8bitmaker’s file9.5 / 16.0 GB14–21 tok/s128KUnified memory
Runs wellDeepSeek-R1-Distill-Llama-8B DeepSeek8BQ8_09.4 / 16.0 GB14–21 tok/s60KUnified memory
Runs wellGemma 4 E4B Google8BQ8_08.6 / 16.0 GB16–24 tok/s125KUnified memory
Runs wellLing-3.0-tiny Ant Group7.9B / 1.3BQ8_0maker’s file8.5 / 16.0 GB85–127 tok/s128KUnified memory
Runs wellQwen2.5-7B-Instruct-1M Alibaba Qwen7.6BQ8_08.5 / 16.0 GB16–24 tok/s145KUnified memory
Runs wellQwen2.5-7B-Instruct Alibaba Qwen7.6BQ8_0maker’s file8.5 / 16.0 GB16–24 tok/s32KUnified memory
Runs wellDeepSeek-R1-Distill-Qwen-7B DeepSeek7.6BQ8_08.5 / 16.0 GB16–24 tok/s32KUnified memory
Runs wellMistral-7B-Instruct-v0.1 Mistral AI7.2BQ8_08.7 / 16.0 GB16–23 tok/s32KUnified memory
Runs wellGemma 4 E2B Google5.1BBF16 (native)maker’s file10.1 / 16.0 GB13–20 tok/s125KUnified memory
Runs wellWeVisOne-VL-4B Tencent4.8BBF16 (native)maker’s file10.6 / 16.0 GB13–19 tok/s46KUnified memory
Runs wellQianfan-OCR Baidu4.7BBF16 (native)maker’s file10.5 / 16.0 GB13–19 tok/s31KUnified memory
Runs wellQwen3.5-4B Alibaba Qwen4.7BBF16 (native)maker’s file9.4 / 16.0 GB14–21 tok/s218KUnified memory
Runs wellQwen3-VL-4B-Instruct Alibaba Qwen4.4BBF16 (native)maker’s file9.9 / 16.0 GB14–20 tok/s51KUnified memory
Runs wellQwen3-4B-Instruct-2507 Alibaba Qwen4BBF16 (native)maker’s file9.1 / 16.0 GB15–22 tok/s57KUnified memory
Runs wellQwen3-4B Alibaba Qwen4BBF16 (native)maker’s file9.1 / 16.0 GB15–22 tok/s128KUnified memory
Runs wellDeepSeek-OCR DeepSeek3.3BBF16 (native)maker’s file7.2 / 16.0 GB49–74 tok/s8KUnified memory
Runs wellGranite-4.0-H-Micro IBM3.2BBF16 (native)maker’s file6.5 / 16.0 GB21–32 tok/s128KUnified memory
Runs wellQwen3-1.7B Alibaba Qwen2BBF16 (native)maker’s file5.2 / 16.0 GB27–41 tok/s40KUnified memory
Runs wellDeepSeek-R1-Distill-Qwen-1.5B DeepSeek1.8BBF16 (native)maker’s file4.0 / 16.0 GB36–54 tok/s125KUnified memory
Runs wellGLM-OCR Zhipu AI1.3BBF16 (native)maker’s file3.5 / 16.0 GB43–64 tok/s128KUnified memory
Runs wellQwen3-0.6B Alibaba Qwen0.8BBF16 (native)maker’s file2.8 / 16.0 GB56–84 tok/s40KUnified memory
Runs wellEmbeddingGemma 2 Google0.7BBF16 (native)maker’s file2.0 / 16.0 GB87–130 tok/s256KUnified memory
Runs wellQwen3-Embedding-0.6B Alibaba Qwen0.6BBF16 (native)maker’s file2.5 / 16.0 GB64–96 tok/s32KUnified memory
Runs wellbge-reranker-v2-m3 BAAI0.6BFP32 (native)maker’s file3.4 / 16.0 GB44–67 tok/s8KUnified memory
Runs wellbge-m3-unsupervised BAAI0.6BFP32 (native)maker’s file3.4 / 16.0 GB44–67 tok/s8KUnified memory
Runs wellbge-large-en-v1.5 BAAI0.3BFP32 (native)maker’s file1.8 / 16.0 GB98–147 tok/s512Unified memory
Runs wellbge-reranker-base BAAI0.3BFP32 (native)maker’s file1.6 / 16.0 GB121–181 tok/s514Unified memory
Runs wellbge-base-en-v1.5 BAAI0.1BFP32 (native)maker’s file0.9 / 16.0 GB299–448 tok/s512Unified memory
Runs wellbge-small-en-v1.5 BAAI0BFP32 (native)maker’s file0.6 / 16.0 GB955–1433 tok/s512Unified memory
Runs, slowlyQwen3.5-27B Alibaba Qwen27.8BQ4_015.8 / 16.0 GB8–13 tok/s11KUnified memory
Runs, slowlyQwen3.6-27B Alibaba Qwen27.8BQ4_015.8 / 16.0 GB8–13 tok/s11KUnified memory
Runs, slowlyQwen3.8-27B Alibaba Qwen27.8BQ4_015.8 / 16.0 GB8–13 tok/s11KUnified memory
Runs, slowlyMagistral Small 1.2 Mistral AI24BIQ4_XS13.7 / 16.0 GB10–15 tok/s22KUnified memory
Runs, slowlyDevstral Small 1.0 Mistral AI23.6BQ4_014.2 / 16.0 GB9–14 tok/s19KUnified memory
Runs, slowlyDevstral Small 1.1 Mistral AI23.6BQ4_014.2 / 16.0 GB9–14 tok/s19KUnified memory
Runs, slowlyMagistral Small 1.0 Mistral AI23.6BQ4_014.2 / 16.0 GB9–14 tok/s19KUnified memory
Runs, slowlyMagistral Small 1.1 Mistral AI23.6BQ4_014.2 / 16.0 GB9–14 tok/s19KUnified memory

Estimates from the makers’ Hugging Face repositories (as of 2026-11-06); filtered views aren’t indexed. How it’s estimated