Guides
Small models drop tool calls; one did not
Local tool calling in Hermes and OpenCode on a 48 GB Mac mini: five models downloaded and deleted in a day, and the one that worked.
7 minAI
Everything we publish on AI: the models we run on our own machines, what the tools send home, the benchmarks we measure ourselves and the claims we check against the code. Start with the investigations, then the guides.
Insights
Is the M5 Max worth it for local inference?
Memory bandwidth, unified memory ceilings and time to first token against our 48 GB M4 Pro, and whether the upgrade earns its price today.
8 min
Insights
What to run locally and what to send out
A decision guide for 2026: the compliance and cost lines that decide it, with a bank's chatbot as the test case.
12 min
Insights
Gemma 4 E4B makes 8 GB of memory useful
What a small model now does on edge hardware, where it fails, and when it beats reaching for a bigger one.
8 min
Insights
NVFP4 or q4_K_M in Ollama?
The two 4-bit builds are nearly the same size on disk. What the newer format changes for quality and speed, and which to pull.
22 min
Insights
Gemma 4 31B or Qwen 3.5 27B on a Mac?
Which fits, which is faster, and which answers better when a Mac holds one large model at a time.
16 min
Insights
MLX, llama.cpp or MetalRT on Apple Silicon?
One quantised model, three engines, one Mac: generation speed, memory and setup, measured rather than argued about.
12 min
Insights
LM Studio or Ollama on a Mac?
Setup, speed, model management and the API each exposes, plus which we kept on our own machine and why.
15 min
Guides
Gemma, Qwen and Llama speeds on an M4 Pro
Tokens per second, memory used, and the point where the machine starts to struggle.
9 min
Guides
How to run Gemma 4 12B with Ollama
The pull command, the memory it needs, the context you can afford, and how fast it really answers.
9 min