Mac Studio, M4 Max, 64GB, refurbished
1TB of storage, 546GB/s of memory bandwidth, and 59GB of unified memory left for a model once macOS has taken its share. It holds 11 of the 22 models we track.
The 16-core CPU, 40-core GPU M4 Max. The 14-core variant of the same chip runs at 410GB/s, so it would be slower; this row is the faster one.
Offers
checked 23 minutes ago- Apple Certified Refurbishedcheapest todayin stockchecked 23 minutes ago$3,049
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
- Won't fit— headroom
- Won't fit— headroom
- Won't fit— headroom
- Won't fit— headroom
- Llama-4-Scout-17B-16EQ3_K_MTight3GB headroomest. 32.9 tok/s
- Llama-3.3-70BQ5_K_MFits5GB headroomest. 7.2 tok/s
- Qwen3-32BQ8_0Fits22GB headroomest. 10.4 tok/s
- Qwen3-30B-A3BQ8_0Fits24GB headroomest. 77.9 tok/s
- Gemma-3-27BBF16Tight2GB headroomest. 6.6 tok/s
- Fits8GB headroomest. 7.4 tok/s
- gpt-oss-20bBF16Fits15GB headroomest. 37.9 tok/s
- Phi-4BF16Fits28GB headroomest. 12.7 tok/s
- Fits41GB headroomest. 22.2 tok/s
- Gemma-3n-E4BBF16Fits44GB headroomest. 44.4 tok/s
- Qwen3-4BBF16Fits49GB 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 | — | Won't fit | — | — |
| MiniMax-M2 | — | Won't fit | — | — |
| Mistral-Large-2411 | — | Won't fit | — | — |
| gpt-oss-120b | — | Won't fit | — | — |
| Llama-4-Scout-17B-16E | Q3_K_M | Tight | 3GB | 32.9 tok/s |
| Llama-3.3-70B | Q5_K_M | Fits | 5GB | 7.2 tok/s |
| Qwen3-32B | Q8_0 | Fits | 22GB | 10.4 tok/s |
| Qwen3-30B-A3B | Q8_0 | Fits | 24GB | 77.9 tok/s |
| Gemma-3-27B | BF16 | Tight | 2GB | 6.6 tok/s |
| Devstral-Small-2507 | BF16 | Fits | 8GB | 7.4 tok/s |
| gpt-oss-20b | BF16 | Fits | 15GB | 37.9 tok/s |
| Phi-4 | BF16 | Fits | 28GB | 12.7 tok/s |
| Qwen3-VL-8B-Instruct | BF16 | Fits | 41GB | 22.2 tok/s |
| Gemma-3n-E4B | BF16 | Fits | 44GB | 44.4 tok/s |
| Qwen3-4B | BF16 | Fits | 49GB | 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.