Articles for people who
ship the stack
Rewrites on React, Node.js, TypeScript and AI. A translated body is shown when it exists. If it does not, you see the English text and a note that says so.
Tagged: MLX
Hy-MT2-7B on Five Local Runtimes: Which One Should Serve Translation on a Mac?
Hy-MT2-7B load-tested on llama.cpp, LM Studio, mlx-lm, Ollama and MLX-Serve on two Mac Studio M5 Ultra machines: throughput, scaling, determinism, memory.
9 min readRunning Local Coding Agents on a Mac With oMLX: Disk-Backed KV Cache Explained
oMLX caches prompt prefixes on SSD so local coding agents stay fast on Apple Silicon. Install it and connect Claude Code, Codex, Cursor and OpenCode to it.
9 min readInside an LLM: Pure Python Inference for Qwen3-0.6B on MLX
A step-by-step build of LLM inference in about 200 lines of Python with MLX: tokenizer, RMSNorm, RoPE, grouped query attention, SwiGLU, KV cache and sampling.
13 min readRunning AI Agents on a MacBook: Why oMLX Beats Raw MLX
Raw MLX needed about 20 seconds for an 8.7k-token agent context on an M1 Max. oMLX cut it to roughly 3 with batching and a tiered KV cache. Numbers and why.
5 min readRun Local LLMs on a Mac With MLX Instead of the Default llama.cpp Engine
When Apple's MLX beats llama.cpp for a local LLM on Apple Silicon, with speed ranges by model size, three ways to switch from Ollama, and pitfalls to avoid.
5 min readIs a Bigger LoRA Worth It? Measure Accuracy Against Cost, Not Vibes
A small harness that treats LoRA recipes as data and reports accuracy, training time, latency and memory, so rank and iteration choices rest on numbers.
5 min readServing a LoRA Fine-Tune Locally: Verify, Fuse and Avoid Silent Failures
Prove a LoRA adapter actually improved a small model, fuse and serve it behind a local OpenAI-compatible API, and catch the failures that return confident wrong output.
12 min read
About these articles
Request a 24h estimate
Need the same stack in a production operator layer? Send the brief — estimate within 24 hours.