Settings

Configure the AI provider used for question generation. Changes are written to your .env file and take effect on the next request.

AI Provider

Choose the AI backend used when generating questions.

Anthropic Claude

active

Model ID to use for generation (e.g. claude-opus-4-8, claude-sonnet-4-6).

OpenAI / Compatible Endpoint

active

Change to use any OpenAI-compatible endpoint: http://localhost:11434/v1 (Ollama), http://localhost:1234/v1 (LM Studio), or a hosted provider such as Together.ai or Groq.

Model name sent in the request body (e.g. llama3, gpt-4o, mistral).

Groq

active

Get a free key at console.groq.com. Free tier: 14,400 req/day, 30 RPM.

Free models: llama-3.3-70b-versatile (best quality) · llama-3.1-8b-instant (faster, higher quota).

Note: Diagram labelling questions generate images via OpenAI (gpt-image-1). If you use Groq for text generation and need diagram images, keep your OpenAI API key filled in under the OpenAI section above.

Request Timeouts

Maximum time to wait for a response from the AI provider (10–300 s).