MLX Studio - Easiest way to run LLM's on your Mac. All in one engine.
vMLX - Use MLX models easily - JANGQ (GGUF for MLX) - Not dependant on mlx_vlm
JANG — GGUF for MLX. YOU MUST USE JANG_Q RUNTIME. Adaptive Mixed-Precision Quantization + Runtime for Apple Silicon
Heterogeneous prefill/decode for DeepSeek-V4-Flash: CUDA prefill (DGX Spark, vLLM) -> Metal decode (Mac Studio, oMLX) over plain 10GbE
Community-driven benchmark suite for MLX inference engines on Apple Silicon
Privacy-first native macOS cleanup, storage analysis, and system monitoring — Agent workspace, local MLX/Ollama runtime discovery with guarded control, and on-device alerts. Review first; files go to Trash by default.
Verified Qwen3.8 Flash Next oMLX Lightning MTP deployment guide for Apple Silicon
ZBoy - A sovereign, local coding agent on macOS.
1.0 Beta · oMLX 0.6.3rc2 Dual-ANE+CPU · GLM-5.3-Flash Abliterated oQ4 + native MTP · 32/32 · ~24 tok/s · Mac Studio M3 Ultra
GLM-5.3-Flash Abliterated oQ4-MTP on oMLX 0.7.0, Mac Studio M5 Ultra: 60.9 tok/s greedy decode (2.6x the M3U pack), ~1.1k tok/s prefill, Lightning MTP fixed depth 2, one-shot setup
Rei Agent. A local-first agent you assemble yourself.
Memory-safe, LAN-accessible OpenAI-compatible server for Gemma 4 12B (4-bit MLX) on a 16 GB M2 Pro Mac mini
Qwen3.8-27B on Apple Silicon via oMLX + DFlash2 block-diffusion speculative decoding — 128k context on a 48 GB Mac (8-bit KV), single-command bootstrap, measured benchmark ladder
Point it at a video, image, or PDF — get structured JSON. uvx vidlizer[mcp]. Runs local (Ollama/gemma4, LM Studio, oMLX) or cloud (OpenRouter). CLI + MCP server for Claude Code, Cursor, and Claude Desktop.
Frontier orchestration runtime for delegating work across harnesses, models, and providers.
Real DuckDB Quack infrastructure for multi-agent Werewolf: containerized player nodes, Quack gateway federation, browser runner, and local/hosted LLM evals.
To associate your repository with the omlx topic, visit your repo's landing page and select "manage topics."