Mac Studio M3 Ultra 96GB
96 GB unified · 819 GB/s
Prices
Every retailer we track in the UK, cheapest first. Buy links may earn us a commission.
No retailer in this market lists it right now.
What it runs
Every model in the catalogue, with the best quant for this hardware. Estimates, not measurements.
| Model | Quant | Fits | Memory | Max context | Decode @128k | Prefill | Good for |
|---|---|---|---|---|---|---|---|
| Qwen 3.8 27B Alibaba Qwen | Q8_0 | fits | 36.5 GB | 256k | 14tok/sest. | 138 tok/s | chat |
| Qwen 3.5 122B-A10B Alibaba Qwen | Q3_K_M | fits | 57.0 GB | 256k | 41tok/sest. | 245 tok/s | chat, coding |
| Qwen 3.6 27B Alibaba Qwen | Q8_0 | fits | 36.1 GB | 256k | 14tok/sest. | 138 tok/s | chat |
| Qwen3 Coder Next Alibaba Qwen | Q6_K | fits | 65.5 GB | 256k | 49tok/sest. | 543 tok/s | chat, coding, agents, long documents |
| Nemotron 3 Super 120B-A12B NVIDIA | UD-Q3_K_M | fits | 60.0 GB | 256k | 44tok/sest. | 344 tok/s | chat, coding, long documents |
| gpt-oss 120B OpenAI | MXFP4 | fits | 64.9 GB | 128k | 40tok/sest. | 363 tok/s | chat, coding, agents, long documents |
| Gemma 4 31B | Q8_0 | fits | 42.5 GB | 256k | 12tok/sest. | 131 tok/s | chat |
| Mistral Small 4 119B Mistral AI | UD-Q3_K_M | fits | 54.8 GB | 745k1024k with q8 cache | 51tok/sest. | 212 tok/s | chat, coding |
| Qwen 3.6 35B-A3B Alibaba Qwen | Q8_0 | fits | 38.3 GB | 256k | 55tok/sest. | 669 tok/s | chat, coding, agents, long documents |
| Qwen 3.5 27B Alibaba Qwen | Q8_0 | fits | 36.1 GB | 256k | 14tok/sest. | 138 tok/s | chat |
| Gemma 4 26B-A4B | Q8_0 | fits | 29.1 GB | 256k | 45tok/sest. | 812 tok/s | chat, coding, agents, long documents |
| GLM 4.5 Air Z.ai | Q2_K | fits | 66.6 GB | 128k | 10tok/sest. | 62 tok/s | chat, coding |
| Qwen3 Next 80B-A3B Alibaba Qwen | Q6_K | fits | 65.5 GB | 256k | 56tok/sest. | 593 tok/s | chat, coding, agents, long documents |
| Qwen 3.5 35B-A3B Alibaba Qwen | Q8_0 | fits | 38.3 GB | 256k | 55tok/sest. | 669 tok/s | chat, coding, agents, long documents |
| GLM 4.7 Flash Z.ai | Q8_0 | fits | 37.6 GB | 198k | 26tok/sest. | 316 tok/s | chat, coding, agents, long documents |
| Devstral Small 2 24B Mistral AI | Q8_0 | fits | 44.7 GB | 296k384k with q8 cache | 11tok/sest. | 123 tok/s | chat |
| Nemotron 3.5 Lightning 30B-A3B NVIDIA | UD-Q6_K_XL | fits | 34.7 GB | 256k | 76tok/sest. | 900 tok/s | chat, coding, agents, long documents |
| Kimi Linear 48B-A3B Moonshot AI | Q8_0 | fits | 51.0 GB | 1024k | 75tok/sest. | 502 tok/s | chat, coding, long documents |
| gpt-oss 20B OpenAI | MXFP4 | fits | 15.6 GB | 128k | 57tok/sest. | 526 tok/s | chat, coding, agents, long documents |
| Qwen3 Coder 30B-A3B Alibaba Qwen | Q8_0 | fits | 43.6 GB | 256k | 20tok/sest. | 195 tok/s | chat, coding, agents |
| Qwen3 32B Alibaba Qwen | Q8_0 | fits | 65.8 GB | 40k | 8tok/sest. | 55 tok/s | – |
| Gemma 4 12B | Q8_0 | fits | 15.5 GB | 256k | 35tok/sest. | 338 tok/s | chat, coding, long documents |
| Qwen 3.5 9B Alibaba Qwen | Q8_0 | fits | 14.3 GB | 256k | 38tok/sest. | 402 tok/s | chat, coding, long documents |
| Gemma 3 27B | Q8_0 | fits | 38.5 GB | 128k | 13tok/sest. | 169 tok/s | chat |
| Mistral Small 3.2 24B Mistral AI | Q8_0 | fits | 44.7 GB | 128k | 11tok/sest. | 123 tok/s | chat |
| DeepSeek R1 Distill Qwen 32B DeepSeek | Q8_0 | fits | 65.8 GB | 128k | 8tok/sest. | 74 tok/s | – |
| Ministral 3 14B Mistral AI | Q8_0 | fits | 34.7 GB | 256k | 15tok/sest. | 158 tok/s | chat, coding |
| Qwen3 14B Alibaba Qwen | Q8_0 | fits | 36.0 GB | 40k | 14tok/sest. | 135 tok/s | chat, coding |
| DeepSeek R1 0528 Qwen3 8B DeepSeek | Q8_0 | fits | 27.5 GB | 128k | 19tok/sest. | 204 tok/s | chat, coding |
| Qwen3 8B Alibaba Qwen | Q8_0 | fits | 27.5 GB | 40k | 19tok/sest. | 204 tok/s | chat, coding |
| Phi-4 14B Microsoft | Q8_0 | fits | 40.9 GB | 16k | 13tok/sest. | 135 tok/s | chat, coding |
| Qwen 3.5 4B Alibaba Qwen | Q8_0 | fits | 9.6 GB | 256k | 60tok/sest. | 625 tok/s | chat, coding, long documents |
| Ministral 3 8B Mistral AI | Q8_0 | fits | 26.8 GB | 256k | 20tok/sest. | 206 tok/s | chat, coding |
| Granite 4.2 8B IBM | Q8_0 | fits | 30.1 GB | 128k | 17tok/sest. | 185 tok/s | chat, coding |
| Gemma 4 E4B | Q8_0 | fits | 12.5 GB | 128k | 44tok/sest. | 563 tok/s | chat, coding, long documents |
| LFM2.5 8B-A1B Liquid AI | Q8_0 | fits | 15.7 GB | 125k | 43tok/sest. | 752 tok/s | chat, coding, long documents |
| Llama 3.1 8B Meta | Q6_K | fits | 23.5 GB | 128k | 22tok/sest. | 222 tok/s | chat, coding |
| DeepSeek V4 Flash DeepSeek | UD-Q2_K_XL | won't run | 102.3 GB | – | – | – | – |
| GLM 5.3 Flash Z.ai | UD-Q2_K_XL | won't run | 104.1 GB | – | – | – | – |
| Qwen 3.8 Flash Next Alibaba Qwen | UD-Q2_K_XL | won't run | 77.9 GB | – | – | – | – |
| MiniMax M2.7 MiniMax | UD-Q2_K_XL | won't run | 102.5 GB | 5k9k with q8 cache | – | – | – |
| Llama 3.3 70B Meta | Q4_K_M | won't run | 81.0 GB | 99k128k with q8 cache | – | – | – |
Specifications
From the manufacturer's published specifications; sources are linked where we have them.
- Chip
- Apple M3 Ultra (80-core GPU)
- Memory bus
- 1024-bit
- Memory bandwidth
- 819 GB/s
- FP16 compute
- 28 TFLOPS
- Software
- Metal
- Unified memory
- 96 GB
- GPU can use by default
- 72 GB
- GPU can use, tuned
- 88 GB
- Operating system
- macos
- Processor
- M3 Ultra 32-core
- Storage
- 1000 GB