Show HN: SlothSpeak: open-source BYO-API-K mobile chat with the best AI models (github.com)

SlothSpeak has launched as an innovative open-source mobile app designed for voice chat using advanced large language models (LLMs). Unlike typical voice applications that prioritize speed, SlothSpeak intentionally opts for response quality, often taking minutes to deliver natural-sounding answers based on powerful models like OpenAI's GPT-5.2 and Gemini 3.1 Pro. This approach allows users to engage in multi-turn conversations without needing to type, enabling hands-free interaction through voice activity detection. The significance of SlothSpeak lies in its unique architecture—without a server backend, the app interacts directly with AI provider APIs from the user's device, ensuring enhanced privacy and lower latency for API key management. The app supports multiple AI models, allows for customizable reading styles, and includes features like context-aware follow-up prompts and in-depth research capabilities. Enhanced by a robust audio playback system and detailed citation tracking, SlothSpeak opens new avenues for utilizing AI in conversational interfaces, addressing a gap in high-quality, interactive voice-driven applications for both casual users and researchers.