🕒 Changelog
Below are the main updates for each version of NextAI.
Release dates are based on HarmonyOS NEXT AppGallery publication.
🚀 Version 1.1.0
Release Date: September 1, 2025
✨ Core Features
- Added multimodal conversation support — upload images (if supported by model) for text-image interactions.
- Native Markdown rendering, with full support for code blocks, tables, and LaTeX formulas.
- Added Token usage estimation, making it easier to manage model costs.
- Added context memory and custom System Prompt, improving conversation continuity and flexibility.
📤 Sharing & Export
- New multi-format export options: Image, PDF, Markdown, and Plain Text.
- Added regenerate response for quick comparison between outputs.
🎙️ Voice & Interaction
- Added speech-to-text for hands-free input.
- Added voice playback for AI responses, with natural TTS output.
- Added typing-feedback vibration, enhancing immersion.
- Supports streaming output, displaying results as they generate.
⚙️ Personalization
- Introduced custom API management, allowing multiple AI service providers (OpenAI, Qwen, DeepSeek, etc.).
- Supports light / dark themes.
- One-tap data cleanup, ensuring privacy and device efficiency.
🛠️ Stability & Optimization
- Improved scroll capture and rendering performance.
- Enhanced app launch speed and memory optimization.
- Fixed several known issues.
🚀 Future Plans
NextAI will continue to evolve based on user feedback. Upcoming features include:
- Model performance comparison and recommendations;
- Local vector retrieval and knowledge base management;
- Enhanced multimodal support (voice / image / document);
- Built-in model templates to reduce manual setup.