eM Client’s current AI setup is locked to ChatGPT through its own paid tier, which brings a few issues. Users have to pay eM Client for AI features even if they already subscribe to ChatGPT, Claude, or other AI services, instead of being able to use the subscriptions they already have. It’s also unclear if eM Client acts as a middleman, possibly causing double charges for those with existing AI subscriptions. On top of that, users are stuck with one AI provider regardless of their preferences, privacy concerns, or investments in other platforms, even though different AIs shine at different tasks and could better suit individual workflows.
I suggest allowing users to choose their preferred AI provider by using their own API keys. Supporting local models would give privacy-conscious users full control over their data. Direct API usage ensures transparency, so users know exactly what they’re paying for. Options could include OpenAI ChatGPT (with the user’s own API key or subscription), local or self-hosted models like Ollama or LM Studio, and other commercial providers.
A settings panel could let users choose their preferred AI backend and enter their own API credentials, while keeping the eM Client-managed ChatGPT integration as a handy default for those who’d rather not deal with API keys. This way, it caters to users who’ve already invested in AI services and presents eM Client as a flexible, user-focused platform instead of a closed system.