Getting a Gemini API Key

Getting a Gemini API key

Castwright can use Google's Gemini models to analyse your book — finding who speaks and tagging each line — and, optionally, to voice it. Google gives every account a free tier that's more than enough to get started; you just need an API key. It takes about a minute.

In a hurry? Go straight to https://aistudio.google.com/apikey, sign in, create a key, copy it, and paste it into Castwright. The steps below walk through exactly that.

Step by step

1. Open the Google AI Studio → API keys page linked above and sign in with any Google account. (First visit? Accept the terms when prompted.)
2. Click Create API key. If you're asked to pick a Google Cloud project, the default one it offers is fine.
3. Your key appears — a long string starting with AIza…. Click Copy.
4. Back in Castwright, go to Account → Server configuration → Gemini API key, paste the key into the field, and click Save key. (The same field appears in the setup wizard's Analysis step, and a "Get a Gemini API key" link beside it brings you back to this page.)

That's it — pick any Gemini model under Defaults for new books → Analysis model and Castwright will use your key.

Good to know

• The free tier is fine for analysis. You don't need billing enabled to start. Google publishes the current free-tier limits on the same site.
• Your key stays on your machine. Castwright stores it in your per-user settings file (~/.castwright/user-settings.json, plaintext, gitignored) — the same trust model as server/.env. It's never sent anywhere but Google.
• Analysis fails immediately with "Gemini … returned an empty response"? That's Google's recitation filter blocking a copyrighted-book chapter, not a bad key. Switch the analysis model to gemma-4-31b-it, or go fully local with Ollama. See Installing Castwright and Analysis & the Analyzer for the details.
• Prefer not to use the cloud at all? A local Ollama model is the free, private alternative — no key required. See Analysis & the Analyzer.