About

Build, run, and own your AI agents

Cruq AI is the platform for teams that want durable AI agents they actually control - not a chatbot, and not a black box. You design agents as graphs of prompts, tools, and model calls, run them in production, and keep every run, trace, and byte of memory.

We started Cruq AI because the fastest-moving teams were stuck between throwaway prototypes and rigid, vendor-locked platforms. Owning your models, your data, and your agents should be the default, not an enterprise add-on. That belief runs through every part of the product.

What we believe

You own the agents

Agents you build on Cruq AI are yours - the graphs, prompts, tools, memory, and run history. Export them, self-host the runtime, or take your data with you. We are infrastructure, not a walled garden.

Bring your own models

Connect your own provider keys (OpenRouter, OpenAI, Anthropic, and more). You pay providers directly at cost, and switch or compare models without re-platforming. No per-credit markup, no lock-in to a single vendor.

Production, not demos

Guardrails, evals, observability, versioned graphs, and full run and tool-call traces are built in - the parts that decide whether an agent survives contact with real users, not just a slick prototype.

Built to be integrated

A tenant-scoped REST API, an MCP server, and OAuth 2.0 mean agents and other software can drive Cruq AI directly. The platform is meant to sit inside your stack, not replace it.

The team

Cruq AI is an independent, founder-led company. We work closely with the teams building on us, ship in the open, and answer our own support. Read more about who is behind it on the team page.

Get in touch

Evaluating a pilot, integrating the API, or rolling agents out across a team? A real person reads every message.