Sealbook
Sealbook / Open source / AI agents & LLM apps

youssofal/MTPLX

The fastest way to run Qwen 3.8 Flash Next, Qwen 3.8 27B and Ternary Bonsai 2 27B on a Mac: 125 tok/s in OpenCode on an M5 Max, and a 27B model on 16 GB Macs. Native MTP speculative decoding on Apple Silicon, exact at any temperature. OpenAI and Anthropic compatible local server.

anthropic-compatibleapple-siliconclaude-codeflash-nextinference-enginellm-inferencelocal-ailocal-llmmacosmetal
GitHub ↗Website ↗

License & maintenance

Apache-2.0 is a permissive license: free to use commercially with attribution.

Last code push 10 days ago.

Similar projects

RepositoryStarsLanguageLicenseActivity
1manaflow-ai/cmuxOpen source Ghostty-based macOS terminal with vertical tabs and notifications for AI coding agents. Built for multitasking, organization, an★ 28.2kSwiftOtherActive
2mozilla-ai/llamafileDistribute and run LLMs with a single file.★ 26.4kC++OtherActive
3jundot/omlxLLM inference server with continuous batching & SSD caching for Apple Silicon — managed from the macOS menu bar★ 22.9kPythonApache-2.0Active
4morluto/reaReverse engineer anything with agents, from app behavior down to native binaries.★ 319.5kTypeScriptMITActive
5affaan-m/ECCThe agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Code★ 283.1kJavaScriptMITActive
6NousResearch/hermes-agentThe agent that grows with you★ 254.9kPythonMITActive
7DietrichGebert/ponytailMakes your AI agent think like the laziest senior dev in the room. The best code is the code you never wrote.★ 166.9kJavaScriptMITActive
8farion1231/cc-switchA cross-platform desktop All-in-One assistant for Claude Code, Codex, OpenCode, OpenClaw, Grok Build & Hermes Agent. Only official website: ★ 145.8kRustMITActive