Ollaya – Ollama for open-source, Jev-style decision models

Ask typed questions concerning any content or JSON and get calibrated answers in milliseconds. Private, open source, on your own hardware.

Fast

Decisions in milliseconds.

A decision example answers in a sole onward pass, alongside no token-by-token generation. On your own GPU, a five-question petition to Laya takes concerning 10 ms, end to end through the HTTP API.

Drop-in compatible

Speaks TypeSafe's API.

Ollaya serves /v1/systemone and /v1/models alongside TypeSafe's petition and reply shapes. The authoritative TypeSafe Python SDK 0.7.1 plant unchanged against a local server.

 export TYPESAFE_BASE_URL=http://localhost:11435 export TYPESAFE_API_KEY=local export TYPESAFE_DEFAULT_MODEL=laya curl http://localhost:11435/v1/systemone -d '{ "model": "laya", "state": "Can I get an invoice for final month?", "questions": { "intent": { "type": "choice", "instructions": "What does the client want?", "criteria": { "invoice": "Needs an invoice or receipt", "refund": "Wants prosperity back", "other": "Anything else" } } } }'

{ "model": "laya:en", "answers": { "intent": { "type": "choice", "choice": "invoice", "confidence": 0.9547, "probabilities": { "invoice": 0.9698, "refund": 0.0172, "other": 0.013 } } }, "usage": { "input_tokens": 43, "output_tokens": 0 } }

Open models

Open weights, prepared to pull.

Start alongside Laya from Convai Innovations: an English model, a 100+ tongue model, a example fine-tuned for typed decisions, and a router that picks for you.

Your data stays yours

Private by default.

Tickets, emails and person messages are frequently the most delicate data you have. With Ollaya they are scored anywhere they already live.

Platforms

Runs anywhere you work.

A desktop app and a command row for macOS, Windows and Linux, and a Docker depiction for servers. Every example runs on the CPU; an NVIDIA GPU on Linux, in WSL 2 or in Docker takes a petition downward to milliseconds.