Skip to main content
Version: 1.42

Okteto AI

Okteto AI lets you launch AI-powered development agents directly from the Okteto Dashboard. Each agent runs in its own isolated Kubernetes namespace, where it can write code, install dependencies, run tests, and debug — all without any local setup on your machine.

You describe a task in natural language, and the agent handles the rest: cloning your repo, setting up the environment, writing and validating code, and producing a pull request when it's done. You can monitor progress in real time through the Dashboard and review changes in an embedded editor.

How it works​

  1. Open the Agents tab in the Okteto Dashboard
  2. Choose a repository or start a new project from scratch
  3. Describe your task — the agent clones, builds, codes, and tests in a sandboxed environment
  4. Review the results — preview live endpoints, inspect code changes, and check test output
  5. Create a PR when you're satisfied with the changes

Each agent gets its own namespace with dedicated compute, filesystem, and network access. Agents don't share state unless you explicitly configure them to.

What agents can do​

  • Build features — new endpoints, services, UI components, or full applications from a description
  • Fix bugs — investigate issues with full access to logs, databases, and runtime state
  • Write tests — generate unit, integration, or end-to-end tests for existing code
  • Refactor code — modernize patterns, improve performance, or migrate to new frameworks
  • Generate documentation — READMEs, API specs, and inline docs based on the actual codebase

Why Okteto AI​

No local setup required. Agents run entirely in the cloud. You don't need Docker, Kubernetes, or any CLI tooling installed on your machine.

Full isolation. Each agent operates in its own Kubernetes namespace. One agent can't affect another, and none of them touch your production environment.

Production-like environments. Agents have access to the same runtime, secrets, and configuration as your real development setup, so what works in the agent's environment works everywhere else.

Parallel execution. Run multiple agents simultaneously on different tasks — one adding a feature, another writing tests, a third fixing a bug — without conflicts.

Built-in observability. Full logs, metrics, and traces for every agent action. You can see exactly what the agent did and why.

Example use cases​

  • Spin up a new service from a prompt like: "Create a new TypeScript REST API with an /alive endpoint and a README"

  • Refactor legacy code with: "Update this repo to use async/await instead of callbacks"

  • Add a feature in parallel to existing work: "Add a banner to the homepage announcing the beta launch"

Getting started​

For administrators​

Before team members can use Okteto AI, an administrator must enable and configure it. See the Admin Configuration Guide for setup instructions.

For developers​

See the Getting Started Guide to launch your first agent.

FAQs​

Can I run multiple agents at once? Yes. Each agent runs in its own namespace and works independently.

Do I need to install anything locally? No. Everything runs in the cloud through the Okteto Dashboard.

Can I bring my own LLM key? Yes. You can provide your own Anthropic API key in the agent settings to use your own quota.

Do you plan to support additional LLM providers? Yes. We plan to add support for additional models based on feedback in a coming release.

Using your own agent tooling​

Okteto AI is a fully managed experience in the Dashboard. If you prefer to use your own AI agent tooling — such as Claude Code in your IDE or agents triggered by CI pipelines — see Agentic Workflows for how to connect agents to Okteto environments using the CLI.

Feedback​

Okteto AI is currently in beta. To request access, visit okteto.com/ai. We welcome your feedback as we continue to improve the experience.