The fastest way to run Qwen 3.8 on a Mac.

MTPLX is a native Mac app and a command line that runs local language models on Apple Silicon with the model's own multi-token prediction (MTP) heads. It runs Qwen 3.8 Flash Next, the 125B mixture of experts, and Qwen 3.8 27B, plus Qwen 3.6, Qwen 3.5 and Gemma 4. The model drafts several tokens ahead of itself, one batched forward pass verifies the draft, and tokens are committed through exact rejection sampling with residual correction. The output distribution is the model's own at any temperature, and decode runs at around twice the speed of plain decoding: measured 1.6x on a 16 GB M4 Mac mini and 2.24x on an M5 Max.

Measured speeds

Every number below was measured on a MacBook Pro M5 Max with 128 GB, fans verified at maximum, sampled at the model's own settings. Conditions and sources for every row, with the raw logs where they exist, are at mtplx.com/benchmarks.

The speed comes with the output unchanged. On every release MTPLX draws a thousand four-token samples from the fast path and a thousand from the plain path at temperature 1, top-p 0.95, top-k 20, and compares the two distributions by token id. On 2.11.3 they match within the plain path's own noise on both Flash Next and the 27B Quality pack. The acceptance rule is the Leviathan and Chen rejection sampling theorem with residual correction, so temperature=0.6, top_p=0.95 behaves exactly like normal decoding, just faster. There is no second draft model eating your RAM, and no greedy shortcut that quietly changes what the model would have said.

How MTPLX compares with mlx-serve, oMLX, LM Studio, Ollama, llama.cpp and mlx-lm, with a version, a machine and a date on every number: mtplx.com/compare.

Get it

The Mac app is the easiest way in. Download the DMG at mtplx.com, drag it to Applications, and the app takes care of everything else: it checks your hardware, recommends a model that actually fits your memory, downloads it, sets up its own Python engine (no Homebrew needed), installs fan control, puts mtplx on your PATH, and then measures your machine to pick the fastest decoding depth.

Recommended for coding: Qwen 3.8 27B Optimized Speed is a 4-bit dynamic quant with great coding speeds and good quality. Its two siblings sit right under it in the app and CLI: Bare Speed (quickest burst chat speeds, lower quality and slower on long coding tasks) and Optimized Quality (8-bit dynamic quant, good coding speeds and perfect quality). On a Mac with 96 GB or more, Qwen 3.8 Flash Next Optimized Speed is the pick. Qwen 3.6 Optimized Speed V2 remains available directly below them.

The CLI on its own:

brew install youssofal/mtplx/mtplx
mtplx start

or python3 -m pip install mtplx if you prefer pip. All releases are listed at mtplx.com/releases.

Requirements: Apple Silicon (M1 or newer), macOS 14+. 16 GB of memory runs the 4B and 9B models comfortably. Qwen 3.8 Optimized Speed is recommended on Macs with 32 GB or more; on M1 and M2 the app and CLI pick its FP16 build (same weights, native precision for those chips) automatically. Both check your Mac before recommending anything.

Qwen 3.8 Flash Next on a Mac

Qwen 3.8 Flash Next is Qwen's 125B-A6B preview of the Qwen4 architecture: a hybrid GatedDeltaNet mixture of experts with Qwen Sparse Attention and a 51B-parameter n-gram table. MTPLX 2.10.0 was the first Apple Silicon backend for the family, and it runs the model's own MTP head as an exact speculative decoder. Two packs, both for Macs with 96 GB of unified memory or more:

• Youssofal/Qwen3.8-Flash-Next-MTPLX-Optimized-Speed: dynamic 4-bit with the sparse-attention projections at 8-bit. The recommended build. 115.1 GB download including the 32 GB n-gram table, about 83 GB resident.
• Youssofal/Qwen3.8-Flash-Next-MTPLX-Bare-Speed: flat 4-bit, the quickest build. 106.3 GB download, about 74 GB resident.

The n-gram table streams from SSD by default, so the weights stay resident and the table does not have to. Context window 262,144 tokens; 261,120-token prompts decode on 2.11.3. Image input works. In the app, pick "Qwen 3.8 Flash-Next Optimized Speed"; from the terminal:

mtplx serve --model Youssofal/Qwen3.8-Flash-Next-MTPLX-Optimized-Speed

Then point OpenCode, Pi, Hermes, Claude Code, Cline, Cursor or anything that speaks the OpenAI or Anthropic API at http://127.0.0.1:8000. The guide with every measured number and its conditions: mtplx.com/models/qwen3.8-flash-next.

Qwen 3.8 27B on a Mac

Qwen 3.8 27B is the coding flagship for Macs with 32 GB or more. MTPLX shipped it on 15 August 2026, the day after Qwen released it. Three packs, each with an FP16 sibling for M1 and M2 that the app and CLI pick automatically:

• Youssofal/Qwen3.8-27B-MTPLX-Optimized-Speed: 4-bit dynamic, 20.4 GB, 23.6 GB peak. The default for coding. Against the bf16 model on a mixed corpus of code, prose and JSON it agrees on 96.0 percent of top-1 tokens with a KL divergence of 0.012.
• Youssofal/Qwen3.8-27B-MTPLX-Bare-Speed: flat 4-bit, 16.0 GB, 17.0 GB peak. Quickest chat speeds.
• Youssofal/Qwen3.8-27B-MTPLX-Optimized-Quality: 8-bit dynamic, 29.4 GB, 32.7 GB peak, for 36 GB or more. 99.3 percent top-1 agreement with bf16, KL 0.0005.

The guide: mtplx.com/models/qwen3.8-27b.

Models and recommended settings

Every model here is an official MTPLX pack on Hugging Face under Youssofal. "Fits" is the memory the pack actually peaks at while serving, next to the smallest Mac the app and CLI will offer it on. The preset column is what MTPLX resolves by itself, so this table is what you get by doing nothing.

Depth 3 is the launch default. mtplx tune --retune measures autoregressive decoding against each depth on your own Mac and saves a shallower one when a shallo