#147 · Primary category： Inference & Local Deploy
mlx-dspark
Up to 4× faster LLM decoding on Apple Silicon, lossless. Native MLX port of DeepSeek's DSpark & z-lab's DFlash speculative decoding — Gemma-4, Qwen3.8, Muse-Glimmer, Nemotron, LFM2.5, Ornith-1.0, ternary Bonsai-27B.
Project last updated：08/27/26
GitHub Stars
620
Forks
50
Contributors
3
License
MIT
Why we included this project
mlx-dspark ports two speculative-decoding drafters, DeepSeek's DSpark and z-lab's DFlash, to Apple Silicon through MLX, and it does so without changing the output: the target model verifies every token, so generation is faster but byte-identical to normal decoding. That lossless guarantee is what makes it worth trying for anyone running agent loops or serving responses where quality drift is not an option. It targets consumer-size models like Gemma-4, Qwen3, Muse-Glimmer, and Nemotron, and you can run it from the CLI, from Python, or through an OpenAI-compatible server that plugs into LM Studio and other local tools. A benchmark mode measures your own machine and picks a draft length instead of assuming one, which helps if you want to tune for your hardware rather than trust a default.
Articles for this project
No articles for this project yet.
To suggest a topic or contribute an article, contact us.
Related projects in this category
ollama
Get up and running with Kimi-K2.6, GLM-5.2, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.
llama.cpp
LLM inference in C/C++
vllm
A high-throughput and memory-efficient inference and serving engine for LLMs
gpt4all
GPT4All: Run Local LLMs on Any Device. Open-source and available for commercial use.
LocalAI
LocalAI is the open-source AI engine. Run any model - LLMs, vision, voice, image, video - on any hardware. No GPU required.