Builds / Apple M5 Max (unified)
MacBook Pro M5 Max 128 GB: Qwen3.8 Flash-Next at 44.4 tok/s (llama.cpp), 64.5 with MTP on oMLX
Source: heretik.io ↗📏 How measured: reported by the original source; RigBase re-checked the numbers at that source
- GPU(s)
- Apple M5 Max (unified)
- VRAM / unified memory (GB)
- 128
- CPU
- Apple M5 Max
- Motherboard
- not stated
- RAM
- 128 GB unified
- PSU
- not stated
- Other
- MacBook Pro; 262K context, BF16 mmproj loaded, 1024 generated tokens, greedy
- Reported total cost
- not stated
- Power
- not stated
Measured speed (generation tokens/sec)
| Model | Quant | Engine | tok/s | Prompt processing (tok/s) | Details |
|---|---|---|---|---|---|
| Qwen3.8 Flash-Next | UD-IQ4_XS | llama.cpp b76b133, no speculative decoding | 44.4 | 1,140 (prompt length not stated) | context: 262K ctx, empty context📏 1024 generated tokens, greedy (llama-bench tg128: 46.3) |
| Qwen3.8 Flash-Next | UD-IQ4_XS + mtp-Qwen3.8-Flash-Next-Q8_0 head | llama.cpp b76b133, MTP (--spec-draft-n-max 2) | 59.3 | not stated | ⚡ spec-decode (MTP)context: 262K ctx, empty context (45.6 tok/s at 245K)📏 1024 generated tokens, greedy |
| Qwen3.8 Flash-Next | oQ4e MLX (4.72 bpw, Jundot/Qwen3.8-Flash-Next-oQ4e-mtp) | oMLX 0.7.1rc1, no speculative decoding | 50 | not stated | context: 262K ctx📏 Source gives 50-58 tok/s; low end used |
| Qwen3.8 Flash-Next | oQ4e MLX (4.72 bpw) with MTP head | oMLX 0.7.1rc1, MTP + n-gram table on SSD | 64.5 | 1,386 (prompt length not stated) | ⚡ spec-decode (MTP)context: 262K ctx, empty context (59.7 tok/s at 245K)📏 1024 generated tokens, greedy |
Prompt processing = prefill speed (how fast the prompt is read) as reported by the source, with the prompt length it was measured at; blank = the source does not report it. est. TTFT @8k is calculated from the measured prefill speed (8,192 ÷ prefill tok/s), shown only when prefill was measured on a prompt of 512+ tokens. It is not a measurement and ignores load time and caching. measured TTFT is the source's own time-to-first-token. Why prefill matters for coding assistants.
heretik.io write-up, 16 Oct 2026.
Source: https://heretik.io/qwen38-flash-next-mtp-macbook/
Blank fields = not stated in the source. We never estimate build data; GPU street prices above are separately sourced and dated.
Rent before you buy: try this model on a cloud GPU for a few dollars first: Runpod (referral link; sign up with Google to get the new-user bonus)