MTPLX

by youssofal · Claude Skill · ★ 2.7k

About MTPLX

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.

anthropic-compatibleapple-siliconclaude-codeflash-nextinference-engineinklinginkling-2-minillm-inferencelocal-ailocal-llm

Quick Facts

Stars2,741
Forks215
LanguagePython
CategoryClaude Skill
LicenseApache-2.0
Quality Score37.3/100
Open Issues82
Last Updated2026-11-22
Created2026-05-02
Platformsclaude-code, python
Est. Tokens~1710k

Compatible Skills

These tools work well together with MTPLX for enhanced workflows:

  • mlx-dspark — semantic(0.58)+complementary+shared_fw(openai)+rare_topics+same_lang+similar_pop+shared_platform (88%)
  • vllm-mlx — semantic(0.46)+complementary+shared_fw(anthropic,openai)+rare_topics+same_lang+similar_pop+shared_platform (86%)
  • Rapid-MLX — semantic(0.41)+shared_fw(anthropic,openai)+rare_topics+same_lang+similar_pop+shared_platform (69%)
  • claude-code-local — semantic(0.43)+shared_fw(anthropic)+rare_topics+same_lang+similar_pop+shared_platform (67%)
  • lm-proxy — semantic(0.15)+complementary+shared_fw(anthropic,openai)+rare_topics+same_lang+shared_platform (61%)

More Claude Skill Tools

Explore other popular claude skill tools:

View all Claude Skill tools →

Popular Python Agent Tools

Frequently Asked Questions

What is MTPLX?

MTPLX is The fastest way to run Inkling 2 Mini, Qwen 3.8 and Ternary Bonsai 2 27B on a Mac: Inkling 2 Mini at up to 128.8 tok/s and Flash Next at 125 tok/s in OpenCode on an M5 Max, and a 27B model on 16 GB Ma. It is categorized as a Claude Skill with 2.7k GitHub stars.

What programming language is MTPLX written in?

MTPLX is primarily written in Python. It covers topics such as anthropic-compatible, apple-silicon, claude-code.

How do I install or use MTPLX?

You can find installation instructions and usage details in the MTPLX GitHub repository at github.com/youssofal/MTPLX. The project has 2.7k stars and 215 forks, indicating an active community.

What license does MTPLX use?

MTPLX is released under the Apache-2.0 license, making it free to use and modify according to the license terms.

View on GitHub → Browse Claude Skill tools