2.3.1

What's Changed

• Fetch models on startup by @DenizOkcu in #150

Full Changelog: 2.2.3...2.3.1

2.2.3

Reintroduces real streaming for Ollama

Full Changelog: 2.2.1...2.2.3

2.2.2

Full Changelog: 2.2.1...2.2.2

2.2.1

fixes changed citations API response from Perplexity

Full Changelog: 2.2.0...2.2.1

2.2.0

v2.2.0 (April 12, 2024) - URL Parameter Refactoring

Service Configuration

• Service-Specific URLs: Each AI service now has its own dedicated URL parameter in both global settings and note frontmatter:
  • openaiUrl for OpenAI API
  • openrouterUrl for OpenRouter.ai
  • ollamaUrl for Ollama
• Automatic Settings Migration: Added migration logic to standardize service URLs for better consistency and compatibility

Feature Enhancements

• 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.

Bug Fixes & Improvements

• 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)

Mobile Enhancements

• Improved Mobile Support: Enhanced Ollama support on mobile with a CORS-friendly request method
• URL Priority Logic: Refined URL priority for fetching available models

Code Maintenance

• Removed Unused Services: Removed deprecated services for a cleaner codebase
• Dependency Updates: Updated project dependencies to latest versions

v2.1.4 (Beta) - Endpoint Handling

Improvements

• 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

v2.1.3 (Beta) - System Commands Update

Enhancements

• 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

Full Changelog: 2.1.2...2.2.0

2.1.5

Full Changelog: 2.1.4...2.1.5

2.1.4

Full Changelog: 2.1.3...2.1.4

2.1.3

What's Changed

• Fix system commands by @DenizOkcu in #144

Full Changelog: 2.1.2...2.1.3

2.1.2

What's Changed

• fixes default front matter collection order by @DenizOkcu in #140
• removes external links from fetching linked content for context by @DenizOkcu in #141

Full Changelog: 2.1.0...2.1.2

2.1.0

What's New in v2.1.0 🚀

• We've added support for OpenRouter.ai as an LLM provider in our Obsidian plugin.
  Set an OpenRouter.ai API key in the settings, and you can access a wide range of models like
  Gemini, DeepSeek, Llama, Perplexity and many more (full list: https://openrouter.ai/models).
• The new command ChatGPT MD: Select Model allows you to select from all available LLMs (openAI, Ollama, OpenRouter.ai)
  and set the current model for your note.

Full Changelog: 2.0.1...2.1.0