Mac Studio, M4 Max, 128GB, refurbished
1TB of storage, 546GB/s of memory bandwidth, and 120GB of unified memory left for a model once macOS has taken its share. It holds 15 of the 22 models we track.
The 16-core CPU, 40-core GPU M4 Max, with twice the storage of the new M5 Max 128GB it undercuts.
Offers
checked 1 hour ago- Apple Certified Refurbishedcheapest todayin stockchecked 1 hour ago$4,409
Refurbished by Apple, with Apple's one-year limited warranty and 14-day returns. Apple says refurbished supply is usually very limited and often runs out, so this row can go dead faster than a new one.
No reseller undercuts Apple Certified Refurbished on this Config, so Apple Certified Refurbished direct is the price to beat.
Timing
Collecting price history — 2 of 90 days. No timing verdict yet.
Models it runs — 8K context
- Won't fit— headroom
- Won't fit— headroom
- Won't fit— headroom
- Won't fit— headroom
- Won't fit— headroom
- Won't fit— headroom
- Won't fit— headroom
- Qwen3-235B-A22BQ3_K_MTight2GB headroomest. 25.4 tok/s
- MiniMax-M2Q3_K_MTight5GB headroomest. 55.9 tok/s
- Fits15GB headroomest. 3.5 tok/s
- gpt-oss-120bQ6_KFits22GB headroomest. 65.3 tok/s
- Tight1GB headroomest. 15.1 tok/s
- Llama-3.3-70BQ8_0Fits42GB headroomest. 4.8 tok/s
- Qwen3-32BBF16Fits53GB headroomest. 5.5 tok/s
- Qwen3-30B-A3BBF16Fits57GB headroomest. 41.4 tok/s
- Gemma-3-27BBF16Fits63GB headroomest. 6.6 tok/s
- Fits69GB headroomest. 7.4 tok/s
- gpt-oss-20bBF16Fits76GB headroomest. 37.9 tok/s
- Phi-4BF16Fits89GB headroomest. 12.7 tok/s
- Fits102GB headroomest. 22.2 tok/s
- Gemma-3n-E4BBF16Fits105GB headroomest. 44.4 tok/s
- Qwen3-4BBF16Fits110GB headroomest. 44.4 tok/s
| Model | Best quant | Fit | Headroom | Est. speed |
|---|---|---|---|---|
| Llama-4-Behemoth | — | Won't fit | — | — |
| Kimi-K2-Instruct | — | Won't fit | — | — |
| DeepSeek-V3 | — | Won't fit | — | — |
| DeepSeek-R1 | — | Won't fit | — | — |
| Qwen3-Coder-480B-A35B-Instruct | — | Won't fit | — | — |
| Llama-4-Maverick-17B-128E | — | Won't fit | — | — |
| GLM-4.6 | — | Won't fit | — | — |
| Qwen3-235B-A22B | Q3_K_M | Tight | 2GB | 25.4 tok/s |
| MiniMax-M2 | Q3_K_M | Tight | 5GB | 55.9 tok/s |
| Mistral-Large-2411 | Q6_K | Fits | 15GB | 3.5 tok/s |
| gpt-oss-120b | Q6_K | Fits | 22GB | 65.3 tok/s |
| Llama-4-Scout-17B-16E | Q8_0 | Tight | 1GB | 15.1 tok/s |
| Llama-3.3-70B | Q8_0 | Fits | 42GB | 4.8 tok/s |
| Qwen3-32B | BF16 | Fits | 53GB | 5.5 tok/s |
| Qwen3-30B-A3B | BF16 | Fits | 57GB | 41.4 tok/s |
| Gemma-3-27B | BF16 | Fits | 63GB | 6.6 tok/s |
| Devstral-Small-2507 | BF16 | Fits | 69GB | 7.4 tok/s |
| gpt-oss-20b | BF16 | Fits | 76GB | 37.9 tok/s |
| Phi-4 | BF16 | Fits | 89GB | 12.7 tok/s |
| Qwen3-VL-8B-Instruct | BF16 | Fits | 102GB | 22.2 tok/s |
| Gemma-3n-E4B | BF16 | Fits | 105GB | 44.4 tok/s |
| Qwen3-4B | BF16 | Fits | 110GB | 44.4 tok/s |
Best quant is the highest-quality one that still loads with headroom. Every speed is an estimate; the formula is here.