Decide in milliseconds, escalate when unsure.
teleon.dev/docs/system1 answers routine decisions with deterministic rules and fast Ollama Cloud models (ollama-cloud/glm-5.3-flash), and sends a small blind share to a stronger check so its accuracy stays measured.
Most decisions finish on the first two rungs, on CPU, in milliseconds.
What it does
Four rungs
Code rules, offline CPU models, hosted small models, then a full model.
Blind canaries
A small share is re-checked blind by a stronger check, to measure accuracy.
Escalate when unsure
Low-confidence items move up a rung with their evidence.
The contract
A decision records the rung that answered and whether it was sampled as a canary.
decision_cascade_work_item/v1
Endpoints
| Method | Path | Access | What it does | Status |
|---|---|---|---|---|
POST | /v1/decide | RING-2 | Get a fast decision | |
GET | /v1/canaries | RING-1 | Read canary accuracy |
Listed paths appear in the service’s published files today. In the one-host plan every path here is served under api.teleon.dev/v1/system1.
Words on this page
- Canary
- A small share of traffic or checks run against a change first, so a problem shows up early and stays small.
- Deterministic check
- A check written as code that gives the same answer every time for the same input, at code speed and cost.