Mac Studio, M4 Max, 128GB, refurbished vs Mac Studio, M5 Ultra, 256GB
Both hold gpt-oss-120b at Q3_K_M. Mac Studio, M5 Ultra, 256GB decodes an est. 241 tok/s against 110, and costs $5,090 more — $38.80 per extra token per second. Worth it if you read the output as it arrives; not if you queue jobs.
- Fit
- Fits
- Fits
- Headroom
- 61GB
- 189GB
- Est. speed
- 110 tok/s
- 241 tok/s
- Cheapest today
- $4,409 at Apple Certified Refurbished
- $9,499 at Apple
- Per GB of memory
- $34
- $37
- Memory
- 128GB unified memory
- 256GB unified memory
- Usable for a model
- 120GB
- 248GB
- Memory bandwidth
- 546GB/s
- 1.2TB/s
- Storage
- 1TB
- 1TB
- Lowest recorded
- $4,409 in 2d
- $9,499 in 3d
Fit, headroom and speed are computed from each vendor’s published bandwidth and the model’s own size at 8K context. Prices are the cheapest listing we could read on each seller’s page. The arithmetic is here.
Where they actually differ
- Qwen3-235B-A22BTightFits
- DeepSeek-V3Won't fitWon't fit
- Llama-3.3-70BFitsFits
- gpt-oss-120bFitsFits
- Mistral-Large-2411FitsFits
- Llama-4-Scout-17B-16EFitsFits
- Qwen3-32BFitsFits
- Gemma-3-27BFitsFits
- gpt-oss-20bFitsFits
- Llama-4-Maverick-17B-128EWon't fitFits
- Llama-4-BehemothWon't fitWon't fit
- Qwen3-Coder-480B-A35B-InstructWon't fitTight
- Kimi-K2-InstructWon't fitWon't fit
- GLM-4.6Won't fitFits
- DeepSeek-R1Won't fitWon't fit
- MiniMax-M2TightFits
- Qwen3-30B-A3BFitsFits
- Qwen3-4BFitsFits
- Devstral-Small-2507FitsFits
- Phi-4FitsFits
- Gemma-3n-E4BFitsFits
- Qwen3-VL-8B-InstructFitsFits
Every model we track at Q3_K_M. The rows where the two boxes agree are dimmed, because they are not what you are choosing between.