The verdict is the new MacBook Pro M5 Max (128 GB) at $6,699, unless a used M3 Max (128 GB) costs under about $4,650. The LLMCheck index puts DeepSeek V4 Flash at 39 tok/s (community run) on it against 27 tok/s (community run) on the used Mac, and both fit the same 93 of the 130 models.
| M3 Max (128 GB) | M5 Max (128 GB) | |
|---|---|---|
| Chip | M3 Max | M5 Max |
| GPU cores | 40-core | 40-core |
| Memory bandwidth | 400 GB/s | 614 GB/s |
| Unified memory | 128 GB | 128 GB |
| Price | — | $6,699 |
| Price per GB | — | $52 |
| Models that fit (of 130) | 93 | 93 |
| Best local LLM here | Qwen3.8-Flash-Next estimated35 tok/s | Qwen 3.6-35B-A3B community130 tok/s |
| DeepSeek V4 Flash | community27 tok/s | community39 tok/s |
| Price per tok/s on DeepSeek V4 Flash | — | $172 |
| Where to buy | Used / refurbished (Apple no longer sells it new) | New · build-to-order from Apple |
Every speed below is the LLMCheck index’s figure for that exact configuration — the same number its MacBook Pro M3 Max (128 GB) and MacBook Pro M5 Max (128 GB) pages print, with its provenance mark. Across the 89 shared models with a figure on both, the M5 Max column is a median ×1.47 of the M3 Max column.
| Model | Size | M3 Max tok/s | M5 Max tok/s | Ratio |
|---|---|---|---|---|
| Qwen3.8-Flash-Next | 125B MoE | estimated35 | community47 | ×1.34 |
| DeepSeek V4 Flash | 284B MoE | community27 | community39 | ×1.44 |
| GLM-5.3-Flash | 320B MoE | estimated17 | estimated27 | ×1.59 |
| DeepSeek V4 Flash Vision-Exp | 305B MoE | estimated24 | estimated39 | ×1.62 |
| Inkling-Small caveat | 276B MoE | estimated29 | estimated42 | ×1.45 |
| KAT-Coder-V2.5 | 35B MoE | estimated84 | community89 | ×1.06 |
| Qwen 3.6-35B-A3B | 35B MoE | estimated92 | community130 | ×1.41 |
| Qwen3.8-27B | 27.8B | estimated19 | community31 | ×1.63 |
| Qwen3-235B-A22B | 235B MoE | estimated17 | estimated26 | ×1.53 |
| Nanbeige4.3-33B | 33.1B | estimated15 | community25 | ×1.67 |
| Recursive Flint | 39.8B MoE | estimated89 | community123 | ×1.38 |
| Ling-3.0-flash-VL | 124B MoE | estimated37 | estimated53 | ×1.43 |
| Qwen 3.6-27B | 27B | estimated20 | estimated30 | ×1.50 |
| Nemotron 3.5 Super VL 120B-A12B | 120B MoE | estimated36 | community36 | ×1.00 |
| Nemotron 3.5 Lightning | 30B MoE | estimated85 | community117 | ×1.38 |
Nothing: both fit the same 93 of the 130 models in the LLMCheck index, because both have 128 GB. The difference is speed and price.
The new M5 Max (128 GB) costs $6,699 (Apple list, 14 Nov 2026): $52 per GB and $172 per tok/s on DeepSeek V4 Flash. A used M3 Max (128 GB) matches that $172 per tok/s at about $4,650; below that price it is the better value on decode speed alone. The index records no used prices: check current listings, and weigh the warranty, battery health and the remaining years of macOS support.
Bottom line: The new M5 Max (128 GB), unless the used one costs under about $4,650.
Renewed stock of the 128 GB MacBook Pro M3 Max varies: check the memory in the listing before you buy. Amazon does not stock the MacBook Pro M5 Max (128 GB): Apple builds it to order.
As an Amazon Associate, LLMCheck earns from qualifying purchases. Affiliate links cost you nothing extra and never influence the verdicts.
Each speed is decode (generation) tok/s from the LLMCheck index for that exact configuration: the lower-median measured run on that chip where one exists (a filled square is a published figure, a hollow circle a community run), otherwise an estimate from the memory-bandwidth model (hollow diamond, printed with a tilde). LLMCheck runs no tests of its own. Prompt processing (prefill) depends on GPU compute rather than bandwidth, and the index does not estimate it; Apple’s M5-generation GPUs add per-core Neural Accelerators aimed at that work.
The verdict is the new MacBook Pro M5 Max (128 GB) at $6,699, unless a used M3 Max (128 GB) costs under about $4,650. The LLMCheck index puts DeepSeek V4 Flash at 39 tok/s (community run) on it against 27 tok/s (community run) on the used Mac, and both fit the same 93 of the 130 models.
Decode speed on Apple Silicon follows memory bandwidth: the MacBook Pro M5 Max (128 GB) reads memory at 614 GB/s, the MacBook Pro M3 Max (128 GB) at 400 GB/s. The LLMCheck index puts DeepSeek V4 Flash at 39 tok/s (community run) against 27 tok/s (community run). Across the 89 shared models with a figure on both, the median gain is 1.47x.
Yes. Both fit the same 93 of the 130 models in the LLMCheck index.
About $4,650 or less. At that price it matches the new Mac's $172 per tok/s on DeepSeek V4 Flash ($6,699 at Apple list, checked 14 Nov 2026). The LLMCheck index records no used prices, so check current listings, and weigh warranty and macOS support.
On the MacBook Pro M3 Max (128 GB): Qwen3.8-Flash-Next at ~35 tok/s (estimated). On the MacBook Pro M5 Max (128 GB): Qwen 3.6-35B-A3B at 130 tok/s (community run). These are the headline picks of the LLMCheck Best by Mac pages: the best-ranked model that reaches 15 tok/s on that Mac and has a standard Mac runtime.