Seven models, every pack shipped with the draft head MTPLX needs for exact speculative decoding on Apple Silicon. The Qwen packs keep the model's own native multi-token-prediction head; the Gemma 4 pack pairs the target with Google's official assistant drafter. Pick by RAM. The app checks your Mac before recommending anything, and auto-tune measures each draft depth on your machine during onboarding, keeps plain AR as the baseline, and saves a depth only if it beats AR. Created by Youssof Altoukhi, who brought native MTP to the Mac in April 2026.
Three packs. Optimized Speed decodes a medium-reasoning coding task at 58.7 tok/s through mtplx serve on an M5 Max, official Qwen 3.8 sampling (temperature 1.0, top-p 0.95, top-k 20), fans verified at max, MTPLX 2.7.0, 15 August 2026. Bare Speed ran 65.2 on the same run.
First Apple Silicon backend for the family (MTPLX 2.10.0, 29 August 2026). Through the MTPLX server on an M5 Max: 61 tok/s plain decode, 63 to 76 tok/s with the MTP head depending on workload, at the pack's own sampler (temperature 1.0, top-p 0.95, top-k 20).
The model MTPLX was built on. Optimized Speed V2 ran 59.9 to 60.1 tok/s on the same medium-reasoning coding task as the Qwen 3.8 packs, mtplx serve, M5 Max, fans verified at max, MTPLX 2.7.0, 15 August 2026.
4-bit MLX body with calibrated INT4 MTP heads. Since MTPLX 2.6.0 (11 August 2026) two agents at once decode at 1.6 to 2.25x per lane against the previous AR batch route on an M5 Max, sampled at shipped settings.
6-bit body, bf16 MTP head. MTPLX 2.0.1 (7 July 2026) took decode from 82.9 to 112.5 tok/s at short context and from 61.6 to 99.7 at 8k on an M5 Max with its 6-bit verify kernels.
2.47 GB download. 227.8 tok/s at depth 3 against 133.6 tok/s plain AR (1.71x) on an M5 Max, fans at max, MTPLX 2.2.0, the card's deterministic suite, 18 July 2026.
Gemma 4 31B IT at MLX 4-bit paired with Google's official Gemma 4 31B assistant drafter at MLX 6-bit, verified with the same exact speculative sampling. In the catalog since MTPLX 1.0.0 (11 June 2026).
Only the tiers MTPLX has published. Where no tier is published the app decides from the machine it runs on, and the memory governor (2.10.0) prints engine budget, weights, resolved context window and session bank in the serve banner. Requests that cannot fit are refused up front with HTTP 507 (2.10.2).
| Unified memory | Run | Published numbers |
|---|---|---|
| 8 GB or more | Qwen 3.5 4B | 2.47 GB download, about 2.9 GiB peak at load. Runs on any Apple Silicon Mac with 8 GB or more. |
| 16 GB | Qwen 3.5 9B or 4B | The README states that 16 GB of memory runs the 4B and 9B models comfortably. |
| 32 GB or more | Qwen 3.8 27B Optimized Speed, the default; Qwen 3.6 27B Optimized Speed V2 | 3.8 Optimized Speed: 20.4 GB download, 23.6 GB measured peak. 3.6 V2: 19.9 GB download. |
| 36 GB or more | Qwen 3.8 27B Optimized Quality | 8-bit. 29.4 GB download, 32.7 GB measured peak. |
| 96 GB or more | Qwen 3.8 Flash-Next | Resident weights about 83 GB (Optimized Speed) or 74 GB (Bare Speed) plus working set; the 32 GB n-gram table streams from SSD. |
| No tier published | Qwen 3.6 35B-A3B, Gemma 4 | The app checks your Mac before recommending anything; auto-tune measures the speed on it. |
M1 and M2 Macs get FP16 siblings of the Qwen 3.8 27B, Qwen 3.6 35B-A3B and Qwen 3.5 9B packs (same weights, native precision for chips without bf16). Every tok/s number on these pages carries its MTPLX version, pack, chip, task, sampler and date; the benchmarks page collects them. Speculation is exact on every pack: drafts are accepted with the probability-ratio rule and rejected drafts are resampled from the residual, so the output follows the model's distribution at any temperature. The acceptance math is on how it works.
Forge converts a Hugging Face repo into an MTP-ready MLX build on your own Mac and measures the speedup; it refuses incompatible models instead of silently falling back. Hugging Face 30-day downloads on 3 September 2026: 61,275 for Qwen 3.8 27B Optimized Speed, 119,875 across the Qwen 3.8 packs, 138,089 across all MTPLX packs. All packs live under huggingface.co/Youssofal.