01 · Context
Threads that remember
Long-running chats plus long-term memory, so you don't re-explain yourself every session.
SQN-01 · Private pilot
An AI workspace for a small crew — documents, live search, long-term memory and an API — running entirely on hardware we own and operate.
Systems
01 · Context
Long-running chats plus long-term memory, so you don't re-explain yourself every session.
02 · Documents
Drop in PDFs, notes and spreadsheets. Answers cite the passages they came from.
03 · Search
Fresh answers with sources, fetched through a search proxy we run. No third-party AI sees your query.
04 · API
OpenAI-compatible /v1 with per-key budgets and rate limits, for scripts and harnesses.
The fleet
Every response is generated on accelerator nodes we own and operate, behind one gateway that handles keys, budgets and usage. Chats and files live on our hardware; backups are encrypted before they leave it.
API
Point any OpenAI-compatible client at api.jmni.ai/v1. Keys are scoped per person, with
their own model list, budget, rate limit and expiry. Stream for anything long.
curl https://api.jmni.ai/v1/chat/completions \ -H "Authorization: Bearer $SQUADRON_KEY" \ -H "Content-Type: application/json" \ -d '{ "model": "deepseek-v4.1-flash", "stream": true, "messages": [{"role": "user", "content": "Status report"}] }'
Squadron is invitation-only while it's in pilot. Tell us who you are and what you'd use it for.