Modelpublic

Youssofal/Qwen3.8-Flash-Next-MTPLX-Bare-Speed

sourceHugging Faceotherupdated 1d agoView on Hugging Face

8likes5.6kdownloads

Model Card

[MTPLX](https://mtplx.com): the fastest way to run Qwen 3.8 on a Mac. Native multi-token-prediction speculative decoding on Apple Silicon, two to three times the speed of plain decoding, exact at any temperature.

Qwen 3.8 Flash-Next Bare Speed

Flat 4-bit quantization. Quickest Flash-Next speeds for chat and coding.

Qwen's 125B-A6B Flash-Next preview, the Qwen4-generation architecture with GDN hybrid MoE, Qwen Sparse Attention, and the 51B-parameter n-gram memory, running natively on MTPLX from day 0, with its native multi-token-prediction head drafting through MTPLX's speculative path. Every expert at flat 4-bit for the fastest Flash-Next build. If you want the higher- quality sibling, pick Optimized Speed.

The 32 GB n-gram table streams from SSD, so only the weights stay in memory. The pack peaks at 78 GiB while serving, which makes it the Flash-Next build for 96 GB Macs.

Measured on MTPLX 2.11.3 (16 September 2026)

This is the Qwen3.8-Flash-Next MLX pack for MTPLX, the fastest way to run Qwen 3.8 Flash Next on a Mac. MacBook Pro M5 Max with 128 GB, fans verified at maximum, sampled at the model's own settings (temperature 1.0, top-p 0.95, top-k 20). Conditions and sources for every row: mtplx.com/benchmarks.

Run	tok/s
One OpenCode request on the Optimized Speed pack: 1,301 tokens generated, 18,539-token prompt with 18,364 tokens served from cache, MTP depth 3	125.8
9k-token code prompt, 1,500 tokens generated, seeded sampler, thinking off (62.5 on MTPLX 2.11.2)	79.3
109k-token OpenCode turn, mean of two runs	61.8
200k-token OpenCode turn, warm, mean of two runs	50.3
Full 45k to 56k-token generations (Flappy Bird at effort xhigh), whole turn	66.8

Exactness on this release: a thousand four-token draws from the fast path match a thousand from the plain path within the plain path's own noise at every joint length, at temperature 1, top-p 0.95, top-k 20. A 96,760-token conversation restored from the session cache in 8 ms. 261,120-token prompts decode. Details: MTPLX 2.11.3 release notes.

Runs on Apple Silicon Macs with 96 GB of unified memory or more: MacBook Pro M4 Max and M5 Max with 128 GB, Mac Studio M3 Ultra, M4 Max and M5 Max. Guide: Run Qwen 3.8 Flash Next on a Mac.

Speeds

Measured on an M5 Max, fans verified at max, single stream, real server (mtplx serve), official Qwen 3.8 sampling (temperature 1.0, top-p 0.95, top-k 20, sampled, not greedy).

Run	tok/s
Coding task, MTP speculative decode (the default)	75.9
Same task, plain autoregressive	47.0

That is a 1.6x speculative multiplier through the product serve path, on sampled output that follows the model's own distribution.

How it is built

—Every MoE expert and dense matrix at 4-bit with 64-weight groups. Nothing promoted, this is the flat, fastest build.
—The GDN convolution and recurrent-state parameters, every norm, the QSA indexer, and the MTP head stay 16-bit.
—The n-gram embedding table ships as a separate ngram-table.safetensors sidecar that MTPLX streams from SSD on every Mac. The vision tower is preserved in the weights.


Download	106.3 GB (includes the 32 GB n-gram table)
Resident weights (n-gram on SSD)	~74 GB + working set
Peak while serving	78 GiB
Recommended Macs	96 GB+ unified memory
Context window	262,144 tokens
MTP depth	adaptive, ceiling 3
Sampling	temperature 1.0, top-p 0.95, top-k 20 (the official Qwen 3.8 contract)

The serving contract ships inside mtplx_runtime.json. MTPLX reads it on load. Drafts are accepted with the probability-ratio rule plus residual resampling, so the output follows the model's own distribution at any temperature.

Use it

Mac app: download at mtplx.com, pick "Qwen 3.8 Flash-Next Bare Speed".

Command line:

bash

pip install mtplx
mtplx serve --model Youssofal/Qwen3.8-Flash-Next-MTPLX-Bare-Speed

Sibling: Optimized Speed (dynamic quant with 8-bit attention, higher quality, slightly slower).

Base model: Qwen/Qwen3.8-Flash-Next (Qwen Community License; the upstream model card is preserved in this repo as README-upstream-qwen.md).