> ## Documentation Index
> Fetch the complete documentation index at: https://docs.corti.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Corti Models

> Access frontier AI models hosted entirely on European infrastructure. Use them in your coding agent or call them directly via API.

Much of today's AI infrastructure runs in US data centers, under US jurisdiction, operated by US companies. For European organizations handling sensitive data, that creates real compliance and sovereignty risks.

**Corti Models is built and hosted entirely in Europe.** Access frontier AI models from your terminal coding agent, or call them directly via API — every prompt is processed on Corti's EU infrastructure.

<CardGroup cols={2}>
  <Card title="European infrastructure" icon="globe">
    All inference and data processing runs on Corti-managed servers in Europe. Your data never leaves the EU.
  </Card>

  <Card title="Frontier performance" icon="trophy">
    The `corti-s1` model family benchmarks above leading proprietary alternatives. See the [model lineup](/models/models) for details.
  </Card>

  <Card title="GDPR compliant" icon="shield">
    No adequacy decisions, SCCs, or legal workarounds. Compliance is built in by default.
  </Card>

  <Card title="Works with your tools" icon="plug">
    An OpenAI-compatible API and a first-class CLI drop straight into the coding agents your team already uses — no rewrites.
  </Card>
</CardGroup>

## Data stays in Europe

All inference, storage, and processing happens on Corti-managed European infrastructure. Your prompts and completions never cross borders, never touch US cloud providers, and are never used to train models. GDPR compliance doesn't require adequacy decisions, standard contractual clauses, or legal workarounds — it is the default.

## Frontier performance, EU-hosted

Choosing European infrastructure doesn't mean compromising on capability. The `corti-s1` model family — available in full-size and compact variants, each with or without reasoning — benchmarks above leading proprietary alternatives, so you get frontier quality on servers that never leave the EU. See the [Models](/models/models) page for the full comparison and pricing.

## Use them your way

There are two ways to build on Corti Models, and both are backed by the same EU-hosted models:

* **In your coding agent** — point OpenCode, ForgeCode, Crush, or Pi at Corti Models with the Corti CLI and start coding in minutes.
* **Directly via API** — the API is OpenAI-compatible, so your existing code works with minimal changes.

## Built for teams that care about compliance

If your industry requires data residency or strict privacy controls, Corti Models removes the compliance workarounds. It runs on the same healthcare-grade platform trusted by large health systems, with enterprise security and certifications inherited out of the box.

## Next steps

<CardGroup cols={2}>
  <Card title="Models" icon="microchip" href="/models/models">
    Compare the four model variants by capability, reasoning, speed, and price.
  </Card>

  <Card title="Quickstart" icon="rocket" href="/models/quickstart">
    Make your first Corti Models API call with any OpenAI SDK.
  </Card>

  <Card title="Code with Corti" icon="terminal" href="/models/ai-coding-agent">
    Connect your terminal coding agent to Corti Models in a few minutes.
  </Card>

  <Card title="API Reference" icon="square-terminal" href="/api-reference/welcome">
    Browse the full Corti Models API with interactive examples.
  </Card>
</CardGroup>