A custom engine and quantization that run a modern 27B model on a 2016 Tesla P100, in one command.
Projects
What I ship now, the older work people still use, and the earlier projects before it. Each one has its own page with the details.
What I ship
The receipts and how to try each one are on its page.
Open-source autoresearch. Agents propose. A judge they can't see or change measures every attempt and keeps only what it confirms.
A lean, local-first core for bots, chats and files. Works with local models and hosted ones. Add only the skills and MCP servers you want.
A source-backed catalog of hosted LLM, embedding, speech, image and research APIs with genuinely free access. Prompt privacy is audited separately from price.
Serves one quantized checkpoint, measures it, and compares it to its full-precision original. Thinking on and off.
Still live, still in use
Not new. Still running, still used every day.
Mandarin Stack is the language-learning workbench I wanted as an English speaker learning Chinese.
SSPS (Stupid Simple Presence Server) uses a single script tag to add presence and visitor tracking to your site. It feels a little like the old hit counters that websites used to have.
An open leaderboard for local and hosted inference. Model, version and engine are measured with one harness held constant, so a difference in scores can be traced to its cause.
Earlier work
Maintained, paused or archived. The pages stay so the history does.
- zd (Zen Suite) A calm, keyboard-first Markdown reader and editor for agent-sized documents. maintained
- Thresher AI-centered security scanner for supply-chain and codebase risk. paused
- Loom ecosystem Small agent and research tools under the larger projects. archived