AI & automation

Configure your LLM (or use none at all)

Bring your own OpenAI-compatible key — or skip it with static sequences or Claude/MCP.

Three ways to power writing

  • Bring your own key — Settings → LLM accepts any OpenAI-compatible endpoint: OpenAI, Groq, Google Gemini, Together, even local Ollama. Your key is encrypted at rest and used only for your generations.
  • No LLM — static sequences with variables and spintax need no model at all.
  • Claude over MCP — Claude writes every email on your subscription via prepare_emails; the platform sends verbatim. Zero LLM config needed.

Setting up a key

  1. 1Go to Dashboard → Settings → LLM.
  2. 2Enter the base URL (e.g. https://api.groq.com/openai/v1) and your API key.
  3. 3Pick a model from the fetched list or type one, then use Test connection to validate.

Groq's free tier with a fast open model is a great starting point if you don't already pay for an LLM.

Still stuck? Ask an AI that can actually fix it.

Open the Sidekick in your dashboard — it has these guides built in and can inspect and fix your account directly.

Or ask Claude with the MCP connector — same guides, same tools, from any Claude app.