Minds uses models from Google, Anthropic, OpenAI and Deepgram, chosen per task for quality, speed and cost. Team admins can pick a preferred AI provider, API callers can override the model per request, and teams can connect their own model endpoint for Mind and Study answers.
Different parts of the platform use different AI models, each picked for the best trade-off between quality, speed and cost. The table below loads live from the models each prompt and code path uses today.
Area
Provider
Model
Mind answers (chats)
Google (default), Anthropic, OpenAI or your own model
gemini-3.8-flash (default) or claude-sonnet-5-5 or gpt-6.1-sol or your own model
Study answers
Google (default), Anthropic, OpenAI or your own model
gemini-3.8-flash (default) or claude-haiku-4-5-20251001 or gpt-6-luna or your own model
