Skip to content
ONTO
Documentation

Install ONTO. Run an agent. Read the full reference.

This page is the on-ramp. The full reference — concepts, SDK, server, models — lives at the canonical docs site, linked below.

Install

Get started in 60 seconds.

No account needed. ONTO defaults to free open-weight models on Ollama Cloud, so you can run your first agent with a single pip install and an env var.

  • No API key relationship needed to start — defaults to free open-weight models.
  • Same primitives in Python, TypeScript, Rust, and over HTTP.
  • On-prem path uses Sled or Postgres + pgvector. No data has to leave your network.
pip install onto-core

Then export OLLAMA_API_KEY to use the free open-weight default. Or set ANTHROPIC_API_KEY / OPENAI_API_KEY to swap providers.

Quickstart — Python

Three lines after install. Defaults to a free open-weight model, so no API key relationship is needed.

examples/python/quickstart.py View on GitHub ↗
from onto import Onto, Models, BuiltinTools

# 1. Initialize. Defaults to Ollama Cloud (free open-weight model).
onto = Onto(
    model=Models.GPT_OSS_20B,
    builtin_tools=BuiltinTools.STANDARD,
)

# 2. Run with consent scopes.
result = onto.run_sync(
    "Hi, I'm Alice. UTC, mornings. Plan me a 3-day trip to Lyon.",
    user_id="alice",
    consent_scopes=["memory:write"],
    extract=True,        # auto-extract typed facts
)

# 3. Inspect what the agent learned.
for ea in result.extracted_assertions:
    print(f"{ea.level:8} {ea.subject}.{ea.predicate} = {ea.object!r}")

# →   profile   alice.timezone = "UTC"
#     profile   alice.prefers  = "mornings"
#     session   alice.location = "Lyon"
Open source

Star the repo, file an issue, send a PR.

ONTO is built in the open. Pre-1.0 we will break the SDK when it makes the SDK better — tracked in the public roadmap.