Skip to main content

Connect Your Own Model Key

On the Free plan, your agents run on your own AI model key. You choose the provider and the model, and you pay the provider directly for what you use. Achilles does not charge you extra for the model.

To make that easy to try, every new Free workspace gets 50 starter messages that run on our model. Your first agent answers straight away. When they run out, you will see a banner reminding you to add a key.

note

Pro, Premium and Enterprise workspaces can run on our keys, or on yours if you prefer. You are never forced to bring a key on a paid plan.

What you need​

An account with one of the supported providers, and an API key from it:

  • Anthropic
  • OpenAI
  • OpenRouter, which reaches many hosted models with one key
  • Sarvam
  • Ollama, cloud or self-hosted

See AI models for how to choose between them.

Add your key​

  1. In the console, open Admin > LLM Providers.
  2. Add a credential for your provider and paste your key. The key is encrypted and is never shown again.
  3. Add the model you want to use from that provider.

Point your agent at it​

  1. Open your agent.
  2. In its model settings, choose the model you just added.

From then on, the agent runs on your key. It no longer uses starter messages, and it does not depend on our model.

If your agent stops answering​

  • "Add your own LLM key" banner or message: your starter messages are used up and the agent is not yet pointed at your model. Follow the steps above.
  • The agent says it cannot reach its model: check that the key is still valid with your provider and that the model name is correct, then see Troubleshooting.

Your keys are yours​

Keys are stored encrypted, scoped to your workspace, and used only to run your agents. See Safety and control.