AI Index

Open-weight models that run on Apple M4 Max (40-core GPU) · 48 GB

  • Unified memory 48 GB
  • Bandwidth 546 GB/s
  • Usable 36.0 GB

Specs: Memory · MacBook Pro (16-inch, 2024) - Tech Specs - Apple Support · Bandwidth · MacBook Pro (16-inch, 2024) - Tech Specs - Apple Support

FitModelParameters / activePrecisionMemoryEst. speedLongest contextRuns as
Runs wellQwen3.5-35B-A3B Alibaba Qwen36B / 3BQ6_K29.0 / 36.0 GB104–156 tok/s256KUnified memory
Runs wellQwen3.6-35B-A3B Alibaba Qwen36B / 3BQ6_K29.0 / 36.0 GB104–156 tok/s256KUnified memory
Runs wellQwen2.5-VL-32B-Instruct Alibaba Qwen33.5BQ4_020.4 / 36.0 GB13–20 tok/s70KUnified memory
Runs wellQwen3-VL-32B-Instruct Alibaba Qwen33.4BQ4_020.3 / 36.0 GB13–20 tok/s70KUnified memory
Runs wellQwen3-VL-32B-Thinking Alibaba Qwen33.4BQ4_020.3 / 36.0 GB13–20 tok/s70KUnified memory
Runs wellQwen2.5-32B-Instruct Alibaba Qwen32.8BQ4_0maker’s file20.2 / 36.0 GB13–20 tok/s32KUnified memory
Runs wellQwen2.5-Coder-32B-Instruct Alibaba Qwen32.8BINT4 (AWQ / GPTQ)maker’s file20.9 / 36.0 GB13–19 tok/s32KUnified memory
Runs wellQwQ-32B Alibaba Qwen32.8BQ4_020.0 / 36.0 GB13–20 tok/s40KUnified memory
Runs wellDeepSeek-R1-Distill-Qwen-32B DeepSeek32.8BQ4_020.0 / 36.0 GB13–20 tok/s32KUnified memory
Runs wellQwen3-32B Alibaba Qwen32.8BMLX 4bitmaker’s file19.0 / 36.0 GB14–21 tok/s128KUnified memory
Runs wellQwen3-Omni-30B-A3B-Captioner Alibaba Qwen31.7BFP831.8 / 36.0 GB51–77 tok/s53KUnified memory
Runs wellGemma 4 31B IT Google31.3BQ4_K_M20.6 / 36.0 GB13–19 tok/s106KUnified memory
Runs wellGLM-4.7-Flash Zhipu AI31.2BFP830.9 / 36.0 GB49–73 tok/s106KUnified memory
Runs wellQwen3-VL-30B-A3B-Instruct Alibaba Qwen31.1BFP8maker’s file32.3 / 36.0 GB56–85 tok/s47KUnified memory
Runs wellQwen3-VL-30B-A3B-Thinking Alibaba Qwen31.1BFP8maker’s file32.3 / 36.0 GB56–85 tok/s47KUnified memory
Runs wellQwen3-30B-A3B-Instruct-2507 Alibaba Qwen30.5B / 3.3BFP8maker’s file31.2 / 36.0 GB65–98 tok/s59KUnified memory
Runs wellQwen3-30B-A3B-Thinking-2507 Alibaba Qwen30.5B / 3.3BFP8maker’s file31.2 / 36.0 GB65–98 tok/s59KUnified memory
Runs wellQwen3-30B-A3B Alibaba Qwen30.5B / 3.3BMLX 8bitmaker’s file31.5 / 36.0 GB65–97 tok/s128KUnified memory
Runs wellQwen3-Coder-30B-A3B-Instruct Alibaba Qwen30.5B / 3.3BFP8maker’s file31.2 / 36.0 GB65–98 tok/s59KUnified memory
Runs wellHy-Memory-30B-A3B Tencent30.1B / 3BQ8_0maker’s file32.0 / 36.0 GB68–102 tok/s50KUnified memory
Runs wellGranite Switch 4.1 30B IBM30BIQ4_XS17.6 / 36.0 GB15–23 tok/s79KUnified memory
Runs wellQwen3.5-27B Alibaba Qwen27.8BQ5_K_M19.8 / 36.0 GB13–20 tok/s256KUnified memory
Runs wellQwen3.6-27B Alibaba Qwen27.8BQ5_K_M19.8 / 36.0 GB13–20 tok/s256KUnified memory
Runs wellQwen3.8-27B Alibaba Qwen27.8BQ5_K_M19.8 / 36.0 GB13–20 tok/s256KUnified memory
Runs wellGemma 4 26B A4B IT Google25.8B / 3.8BQ8_027.3 / 36.0 GB60–89 tok/s230KUnified memory
Runs wellMagistral Small 1.2 Mistral AI24BQ6_K20.5 / 36.0 GB13–19 tok/s107KUnified memory
Runs wellDevstral Small 1.0 Mistral AI23.6BQ6_K20.2 / 36.0 GB13–20 tok/s109KUnified memory
Runs wellDevstral Small 1.1 Mistral AI23.6BQ6_K20.2 / 36.0 GB13–20 tok/s109KUnified memory
Runs wellMagistral Small 1.0 Mistral AI23.6BQ6_K20.2 / 36.0 GB13–20 tok/s39KUnified memory
Runs wellMagistral Small 1.1 Mistral AI23.6BQ6_K20.2 / 36.0 GB13–20 tok/s39KUnified memory
Runs wellERNIE-4.5-21B-A3B-Thinking Baidu21.8B / 3BQ8_023.1 / 36.0 GB75–112 tok/s128KUnified memory
Runs wellGPT OSS 20B OpenAI20.9B / 3.6BMXFP4 (native)maker’s file13.6 / 36.0 GB106–159 tok/s128KUnified memory
Runs wellQwen2.5-14B-Instruct-1M Alibaba Qwen14.8BQ8_016.8 / 36.0 GB16–24 tok/s110KUnified memory
Runs wellQwen2.5-14B-Instruct Alibaba Qwen14.8BQ8_0maker’s file16.9 / 36.0 GB16–24 tok/s32KUnified memory
Runs wellDeepSeek-R1-Distill-Qwen-14B DeepSeek14.8BQ8_016.8 / 36.0 GB16–24 tok/s32KUnified memory
Runs wellQwen3-14B Alibaba Qwen14.8BMLX 8bitmaker’s file16.1 / 36.0 GB16–25 tok/s128KUnified memory
Runs wellGemma 3 12B IT Google12.2BQ8_015.5 / 36.0 GB17–26 tok/s66KUnified memory
Runs wellQwen2.5-Omni-7B Alibaba Qwen10.7BQ8_011.6 / 36.0 GB23–34 tok/s32KUnified memory
Runs wellGLM-4.6V-Flash Zhipu AI10.3BBF16 (native)maker’s file20.4 / 36.0 GB13–20 tok/s125KUnified memory
Runs wellQwen3.5-9B Alibaba Qwen9.7BBF16 (native)maker’s file19.1 / 36.0 GB14–21 tok/s256KUnified memory
Runs wellQianfan-VL-8B Baidu8.8BBF16 (native)maker’s file18.2 / 36.0 GB15–22 tok/s32KUnified memory
Runs wellGranite Switch 4.1 8B IBM8.8BBF16 (native)maker’s file18.5 / 36.0 GB14–22 tok/s115KUnified memory
Runs wellQwen3-VL-8B-Instruct Alibaba Qwen8.8BBF16 (native)maker’s file18.3 / 36.0 GB15–22 tok/s128KUnified memory
Runs wellQwen3-VL-8B-Thinking Alibaba Qwen8.8BBF16 (native)maker’s file18.3 / 36.0 GB15–22 tok/s128KUnified memory
Runs wellpplx-embed-v2-context-9b Perplexity8.4BQ8_09.1 / 36.0 GB30–45 tok/s256KUnified memory
Runs wellQwen2.5-VL-7B-Instruct Alibaba Qwen8.3BBF16 (native)maker’s file16.7 / 36.0 GB16–24 tok/s125KUnified memory
Runs wellQwen3-8B Alibaba Qwen8.2BBF16 (native)maker’s file17.1 / 36.0 GB16–23 tok/s128KUnified memory
Runs wellDeepSeek-R1-Distill-Llama-8B DeepSeek8BBF16 (native)maker’s file16.7 / 36.0 GB16–24 tok/s125KUnified memory
Runs wellGemma 4 E4B Google8BBF16 (native)maker’s file15.8 / 36.0 GB17–25 tok/s125KUnified memory
Runs wellLing-3.0-tiny Ant Group7.9B / 1.3BBF16 (native)maker’s file15.6 / 36.0 GB97–145 tok/s128KUnified memory
Runs wellQwen2.5-7B-Instruct-1M Alibaba Qwen7.6BBF16 (native)maker’s file15.3 / 36.0 GB17–26 tok/s386KUnified memory
Runs wellQwen2.5-7B-Instruct Alibaba Qwen7.6BBF16 (native)maker’s file15.3 / 36.0 GB17–26 tok/s32KUnified memory
Runs wellDeepSeek-R1-Distill-Qwen-7B DeepSeek7.6BBF16 (native)maker’s file15.3 / 36.0 GB17–26 tok/s32KUnified memory
Runs wellMistral-7B-Instruct-v0.1 Mistral AI7.2BBF16 (native)maker’s file15.2 / 36.0 GB18–26 tok/s32KUnified memory
Runs wellGemma 4 E2B Google5.1BBF16 (native)maker’s file10.1 / 36.0 GB26–40 tok/s125KUnified memory
Runs wellWeVisOne-VL-4B Tencent4.8BBF16 (native)maker’s file10.6 / 36.0 GB25–38 tok/s188KUnified memory
Runs wellQianfan-OCR Baidu4.7BBF16 (native)maker’s file10.5 / 36.0 GB26–38 tok/s31KUnified memory
Runs wellQwen3.5-4B Alibaba Qwen4.7BBF16 (native)maker’s file9.4 / 36.0 GB28–43 tok/s256KUnified memory
Runs wellQwen3-VL-4B-Instruct Alibaba Qwen4.4BBF16 (native)maker’s file9.9 / 36.0 GB27–41 tok/s194KUnified memory
Runs wellQwen3-4B-Instruct-2507 Alibaba Qwen4BBF16 (native)maker’s file9.1 / 36.0 GB30–44 tok/s199KUnified memory
Runs wellQwen3-4B Alibaba Qwen4BBF16 (native)maker’s file9.1 / 36.0 GB30–44 tok/s128KUnified memory
Runs wellDeepSeek-OCR DeepSeek3.3BBF16 (native)maker’s file7.2 / 36.0 GB98–147 tok/s8KUnified memory
Runs wellGranite-4.0-H-Micro IBM3.2BBF16 (native)maker’s file6.5 / 36.0 GB42–63 tok/s128KUnified memory
Runs wellQwen3-1.7B Alibaba Qwen2BBF16 (native)maker’s file5.2 / 36.0 GB55–82 tok/s40KUnified memory
Runs wellDeepSeek-R1-Distill-Qwen-1.5B DeepSeek1.8BBF16 (native)maker’s file4.0 / 36.0 GB72–108 tok/s125KUnified memory
Runs wellGLM-OCR Zhipu AI1.3BBF16 (native)maker’s file3.5 / 36.0 GB86–128 tok/s128KUnified memory
Runs wellQwen3-0.6B Alibaba Qwen0.8BBF16 (native)maker’s file2.8 / 36.0 GB112–168 tok/s40KUnified memory
Runs wellEmbeddingGemma 2 Google0.7BBF16 (native)maker’s file2.0 / 36.0 GB173–260 tok/s256KUnified memory
Runs wellQwen3-Embedding-0.6B Alibaba Qwen0.6BBF16 (native)maker’s file2.5 / 36.0 GB128–192 tok/s32KUnified memory
Runs wellbge-reranker-v2-m3 BAAI0.6BFP32 (native)maker’s file3.4 / 36.0 GB89–133 tok/s8KUnified memory
Runs wellbge-m3-unsupervised BAAI0.6BFP32 (native)maker’s file3.4 / 36.0 GB89–133 tok/s8KUnified memory
Runs wellbge-large-en-v1.5 BAAI0.3BFP32 (native)maker’s file1.8 / 36.0 GB196–294 tok/s512Unified memory
Runs wellbge-reranker-base BAAI0.3BFP32 (native)maker’s file1.6 / 36.0 GB241–362 tok/s514Unified memory
Runs wellbge-base-en-v1.5 BAAI0.1BFP32 (native)maker’s file0.9 / 36.0 GB598–896 tok/s512Unified memory
Runs wellbge-small-en-v1.5 BAAI0BFP32 (native)maker’s file0.6 / 36.0 GB1910–2866 tok/s512Unified memory

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