Frontier intelligence, Canadian values

DipoleML exists because Canada should be a place that creates frontier intelligence, not just consumes it. Today we launched the Raven model family and DCode from Canadian soil, and we are setting out the values they are built on.

Aligned with AI for all

Canada's federal AI agenda, from the Pan-Canadian AI Strategy to today's AI-for-all commitments, rests on three ideas: that the benefits of AI should reach everyone, that a sovereign country needs sovereign capability, and that public trust is earned by how you handle data. We built our products around the same three ideas.

• Access. A real free tier on Raven, pay-as-you-go credits with no subscription required, and pricing designed so a student, a startup, and a ministry can all reach the frontier.
• Sovereignty. A Canadian company serving Canadian and global customers from infrastructure we operate. Your models, your billing, and your support should not all live in someone else's jurisdiction.
• Trust. We meter what it takes to run the service and we do not log or store the content of your prompts and completions. PIPEDA and GDPR rights are written into our privacy policy, not bolted on.

How we started

DipoleML began in 2022 as a research habit: a self-taught Canadian builder reading papers, running experiments, and refusing to accept that frontier work only happens south of the border. Four years later that habit is a company with a model family, a coding environment, and a launch date: October 1, 2026.

What we owe the country we build in

A frontier AI company in Canada should make the field more accessible, not less. That means documentation that teaches, pricing that starts at zero, and research published in plain language. It means hiring and partnering here as we grow. And it means building products that treat Canadian values as engineering requirements, not marketing.

The poles have shifted. Intelligence has a northern address. Meet Raven.