- Service-Specific URLs: Each AI service now has its own dedicated URL parameter in both global settings and note frontmatter:
openaiUrl
for OpenAI APIopenrouterUrl
for OpenRouter.aiollamaUrl
for Ollama
- Automatic Settings Migration: Added migration logic to standardize service URLs for better consistency and compatibility
- Perplexity Source Citations: Added support for web source citations when using Perplexity models via OpenRouter.ai. Access models like openrouter@perplexity/llama-3.1-sonar-small-128k-online and openrouter@perplexity/llama-3.1-sonar-large-128k-online without needing a Perplexity Pro subscription.
- Fixed Ollama Streaming: Resolved CORS issues when using Ollama on mobile devices with an alternative request method
- Better Error Handling: Improved feedback for service type and API key validation issues
- System Commands Fix: Fixed missing system commands from notes' frontmatter
- Consistent API Endpoints: Simplified URL and endpoint usage across all services
- Template Organization: Templates now appear in alphabetical order in the template suggest modal (fixes #91)
- Improved Mobile Support: Enhanced Ollama support on mobile with a CORS-friendly request method
- URL Priority Logic: Refined URL priority for fetching available models
- Removed Unused Services: Removed deprecated services for a cleaner codebase
- Dependency Updates: Updated project dependencies to latest versions
- Simplified URL and Endpoint Usage: Streamlined the way service endpoints are handled across the application
- Model Fetching Priority: Improved URL priority for fetching available models
- System Command Handling: Fixed system commands from settings not being applied correctly
- Code Cleanup: Removed unused services and improved code organization
- Template Improvements: Added alphabetical ordering to templates in the template suggestion modal
- Settings Framework: Introduced a robust settings migration framework
- Citation Support: Added support for Perplexity response citations
- Bug Fixes: Fixed various issues reported by users after the 2.1.0 release
- Performance Improvements: Enhanced overall stability and performance
- OpenRouter.ai Support: Added support for OpenRouter.ai as an AI model provider
- Access to models including Gemini, DeepSeek, Llama, Perplexity and more
- Full list of models available at https://openrouter.ai/models
- Model Selection Command: Added new command
ChatGPT MD: Select Model
to choose from all available LLMs (OpenAI, Ollama, OpenRouter.ai) and set the current model for your note
- Service Pattern: Implemented a cleaner service pattern for settings management
- Heading Improvements: Enhanced heading prefix and default frontmatter
- Settings Tab Enhancement: Added API Keys and default URLs to the settings tab for easier configuration
- Version Management: Added comprehensive version management including beta versions
- Title Inference Fix: Prevented error messages from becoming note titles
- API Communication: Simplified communication between services
- Model Discovery: Added a command to list all available models across services
- Stream Management: Improved stream handling and responsiveness
- Clear Chat Fix: Fixed the clear chat command when frontmatter is missing
- Dependency Updates: Updated all dependencies to latest versions
- Service Architecture: Complete refactoring to use service-based architecture
- Multiple AI Providers: Support for OpenAI, Ollama, and OpenRouter
- Enhanced Error Handling: Better error messages and recovery mechanisms
- Context Links: Added ability to include content from linked notes
- Model Name Display: Added model name to assistant divider for better tracking
- Documentation: Completely rewritten README with clearer instructions
- Editor Service: Improved editor service for better text handling
- Dependency Updates: Updated all dependencies to latest versions
- Context Links: Inline link content into messages to the API for better context
- Chat Delimiters: Removed chat delimiters from inlined messages
- Centralized Updates: Streamlined statusBar updates and improved response processing
- Constants: Added command IDs and other important strings as constants
- Service Migration: Moved helper methods to EditorService and OpenAiService
- Code Organization: Improved code structure and imports
- Configuration: Added more configuration values
- Model Extraction: Extracted data structures to Models
- UI Components: Moved ChatTemplatesSuggestModal and Settings to Views
- Frontmatter: Fixed missing frontmatter overwrite with default settings
- Title Inference: Added check for chat folder before renaming inferred title
- Cherry-picked Fixes: Included missing fixes from other branches
- Dependencies: Upgraded all dependencies
- Code Style: Added .prettierrc configuration
- Build Process: Fixed TypeScript build issues
- ChatGPT Comments: Added functionality to ignore parts of your notes using comment blocks
- Conversation Cleanup: Added command to clear conversation except frontmatter
- Language Setting: Added setting for inferring title language
- Source Mode: New chat files now open in source mode for immediate editing
- Template Creation: Fixed new chat from template when
chats
folder is missing - Title Inference: Fixed infer title error that was causing issues
- Stop Button: Added stop button for streaming
- Custom URLs: Added configurable URL support for different API endpoints
- Stream Issues: Fixed stream passthrough issues
- Promise Handling: Fixed resolve/reject handling
- Mobile Experience: Improved streaming icon behavior on mobile devices
- Message Headings: Added headings to messages for better organization
- Code Blocks: Enhanced code block handling for better syntax highlighting
- Streaming: Implemented improved streaming functionality for real-time responses
- SSE Logic: Enhanced error logic for Server-Sent Events
- Title Inference: Added title inference functionality
- Platform Logic: Added platform-specific logic for better cross-platform support
- Custom Dates: Implemented custom date formatting
- Streaming: Fixed streaming bugs for smoother experience
- Title Inference: Reduced message count required for title inference
- Core Functionality: Implemented the foundation of ChatGPT MD
- Error Reporting: Added comprehensive error reporting
- iOS Support: Added compatibility fixes for iOS
- Status Bar: Implemented status bar improvements
- README: Added comprehensive README with usage instructions
- Tutorial: Added YouTube tutorial mirror
- License: Added licensing information
For detailed information about each release, please visit the GitHub repository.