Independent AI Research
Real benchmarks.
No hype.
We run local and cloud AI models head-to-head — speed, accuracy, hallucination resistance, long context — and publish the raw results.
Featured
Benchmarks 8 min read
Gemma4 vs GPT-4o-mini vs Gemini 2.5 Flash: Full Benchmark Results
We ran 5 AI models through 16 tests covering reasoning, hallucination resistance, coding, and long-context recall. Here are the raw results.
April 11, 2026
Recent Articles
Hardware 6 min read
Why Your Mac's 273 GB/s Memory Bandwidth Is the Secret to Fast AI Inference
LLM speed isn't limited by compute — it's limited by how fast you can move model weights from memory to the processor. Here's the physics.
Apr 10, 2026
Tutorials 5 min read
Run Any AI Model Locally on Mac in 10 Minutes (Ollama + Open WebUI)
A complete setup guide for running Gemma4, Llama3, Mistral, and other models locally on Apple Silicon — with a ChatGPT-style web interface.
Apr 9, 2026
Stay in the loop
New benchmark results and AI analysis, straight to your inbox. No spam.
Join the readers. Unsubscribe anytime.