Change Log
Enconvo 2.1.5 Changelog (2025-03-05) π
New Features πβ
-
Added support for new Perplexity models:
r1-1776
andsonar-deep-research
-
Added configurable result count for AI Web Search
-
Added ability to hide Chat Window using the
ESC
key -
Added voice playback functionality for existing responses
-
Added Elevenlabs Scribe speech-to-text integration
-
Added Elevenlabs TTS integration
-
Added Speechify TTS integration
Improvements π§β
- Optimized Live Captions with customizable caption height
- Enhanced stability when adding large files to knowledge base
Bug Fixes πβ
- Fixed longPress shortcut key issue with
Voice input method push
Enconvo 2.1.4 Changelog (2025-03-01) π
New Features πβ
- Added SmartBar default command setting functionality
Allows users to configure preferred commands for SmartBar
- Added
reasoning_effort
parameter for LLMEnables adjustment of model reasoning depth for more control over AI responses
Improvements π§β
- Optimized PopBar hotkey logic to toggle display/hide mode
Improved user experience with more intuitive keyboard shortcut behavior
- Enhanced SmartBar UX logic
Better user interaction flow and responsiveness
Bug Fixes πβ
- Fixed issue where "New Chat" in Chat Window couldn't use web browsing functionality
Restored proper web access capabilities in new chat sessions
- Fixed problem with SmartBar hotkey deletion not working
Ensured reliable hotkey removal when needed
Enconvo 2.1.3 Changelog (2025-02-25) π
New Features πβ
- Added support for the new
Claude 3.7 Sonnet
andClaude 3.7 Sonnet Thinking
models - New
Local Link Reader
provider for faster, more stable, and higher quality webpage content extractionFree and unlimited access to fetch web content without API Keys.
- New
Read Aloud Webpage
Tool for reading aloud the content of a webpage
Improvements π§β
- Optimized prompts to improve agent instruction following capabilities
- Enhanced MCP (Model Context Protocol) stability for more reliable tool execution
Enconvo 2.1.2 Changelog (2025-02-22) π
New Features πβ
- New
Online Video Downloader
tool, supporting platforms (TikTok, YouTube, Twitter, Instagram, Reddit, Vimeo, Pornhub, XVideos, etc.)
You can download videos via EnConvo by simply Ctrl+D on the webpage.
- New
App Context Awareness
feature to use current app content and screenshots as context - New
post-action
settings , which can be executed after command execution, supportingpaste
,insert
andcopy
- New
PopBar
display style settings to show/hide icons and text
Improvements π§β
- Enhanced AI Web Search functionality by adding source references to output results
- Improved Chat Window's
new chat
andclear conversations
logic for better usability - Upgraded Speech-to-text Deepgram Provider with
Nova 3
model support - Optimized
DeepSeek R1
model speed and stability - Added support for custom Google Gemini LLM Provider BaseUrl
- Added support for custom DeepSeek LLM Provider BaseUrl
Bug Fixes πβ
- Fixed auto duplicate issue when using SmartBar in Preview.app
- Fixed text input issues with Voice Input Method when using Local Whisper and Stream Insert
- Fixed Claude image format compatibility issues
- Fixed Insert Below compatibility issues
- Fixed model display not updating in top-right corner after switching models with "/"
Enconvo 2.1.0 Changelog (2025-02-13) π
New Events πβ
- DeepSeek R1 and DeepSeek V3 are now available for unlimited free use (2025-02-13 ~ 2025-03-13)
Improvements π§β
- Optimized default configuration settings logic
Bug Fixes πβ
- Fixed shortcut key settings not taking effect
- Fixed issue where models could not be selected via mouse click
Enconvo 2.0.9 Changelog (2025-02-11) π
New Features πβ
- Added model switching with / character
- Added Context deletion features
- Delete: Remove single Context
- Shift + Delete: Remove single Context but keep input text
- Shift + Cmd + Delete: Remove all Contexts
- Deep Research Agent now shows research process
- Added custom shortcut keys support for @ (search commands) and / (search LLM models) trigger characters
Improvements π§β
- Improved startup stability
- Enhanced Context display
- Improved copy/paste compatibility
Enconvo 2.0.8 Changelog (2025-02-10) π
New Features πβ
- Added Recraft integration for enhanced design workflows
- Added Image Editor Agent for advanced image manipulation
- Added New Chat functionality
- Added Duplicate Command functionality
- Recraft - Crisp Upscale
- Recraft - Generate Image
- Recraft - Image to Image
- Recraft - Vectorize Image
- Recraft - Creative Upscale
- Recraft - Remove Background
- Recraft - Replace Background
Bug Fixes π§β
- Fixed Exa Websearch Provider issues
- Fixed incorrect tab detection in Safari
Enconvo 2.0.7 Changelog (2025-02-07) π
New Features πβ
- Added Deep Research Agent
Feature Updates πβ
- Changed default LLM model to gpt-4o
Bug Fixes π§β
- Fixed background crash issues
- Added multi-screen centered display support for Smart Bar
Enconvo 2.0.6 Changelog (2025-02-06) π
Google Gemini Updates π€β
- Updated Gemini 2.0 Flash to stable release
- Added gemini-2.0-flash-lite model
- Added gemini-2.0-pro-exp model
Subscription Management π³β
- Added subscription cancellation management interface for subscription plan
Bug Fixes π§β
- Fixed Hailuo TTS (Chinese) related issues
- Fixed missing text issue when activating PopBar with hotkey
- Fixed Double Click hotkey configuration issue
- Fixed Voice Input Method unable to stop when connection is lost
EnConvo 2.0.5 Changelog (2025-02-03) π
Video Downloaders π₯β
- Twitter Video Downloader: Added support for downloading Twitter videos
- YouTube Video Downloader: Added support for downloading YouTube videos
Apple Extensions Enhancements πβ
Apple Remindersβ
- Added new commands:
- Update reminders
- Delete reminders
- Toggle reminder status
- Get all incomplete reminders
Apple Calendarβ
- Added new commands:
- Delete events
- Update events
- Get calendar events list
New Apple Mail Extensionβ
- View unread emails
- Delete emails
- Toggle read/unread status
- View email content
- Send emails
Screenshot Extension Updates πΈβ
- Added new capture commands:
- Capture current application
- Capture current screen
- Manual window selection capture
AI Model Improvements π§ β
- Optimized thinking process display for Ollama R1 and other reasoning models
- Added DeepSeek R1 and DeepSeek V3 model support to Enconvo Cloud Plan
- Enhanced Jarvis prompt for improved Agent task execution capabilities
- Updated Perplexity models:
sonar
,sonar-pro
,sonar-reasoning
Other Updates πβ
- Conch TTS: Split into chinese and international servers
- LM Studio: Added model list display
- Fixed straico-related issues
EnConvo 2.0.4 Changelog (2025-01-24) π
Agent Features π€β
-
Jarvis: Jarvis can now automatically complete tasks using tools based on your needs. Default available tools include:
- Code Execution:
- AppleScript: Execute AppleScript automation tasks
- Shell: Run Shell commands and scripts
- Python: Execute Python code and scripts
- NodeJS: Run NodeJS code
- Image Generation: AI-assisted image creation
- Apple Calendar: Manage and create calendar events
- Apple Reminders: Set up and manage reminders
- Image Compression: Optimize image size and quality
- HTTP Requests: Send various network requests
- Text-to-Speech: Convert text to speech
- TTS: Text-to-speech conversion
- Audio/Video Transcription: Convert audio and video to text
- Video Compression: Optimize video file size
- Web Search: Search internet information
- Link Reader: Quick reading and summarizing webpage content
- YouTube Loader: Download and process YouTube video content
- File System Operations: File and directory management
- Code Execution:
-
Added "Create New Agent" feature to create new agents, specify prompts, and configure tools
-
Agent functionality requires models that support Tool Use, Claude 3.5 Sonnet model recommended
MCP (Model Context Protocol) Support πβ
- EnConvo as MCP Client supports installation and use of any MCP Server, with direct server calls in Agents and Workflows
- MCP Settings page features:
- Real-time server status monitoring
- Server configuration modification (changes apply automatically)
DeepSeek r1 Model Support π§ β
- Added support for DeepSeek r1 model
- Added Thinking interaction mode to display model thought process in real-time
New Video Utils Plugin π₯β
- Video Compress: Added video file compression functionality
- Extract Audio: Added audio extraction from video
New Audio Utils Plugin π΅β
- Audio Compress: Added audio file compression functionality
New Transcription Plugin πβ
- Transcribe Audio/Video Files: Added transcription support for audio/video files
Code Runner Plugin Enhancements π»β
- Python Code Runner: Added Python Code Runner support
- NodeJS Code Runner: Added NodeJS Code Runner support
- Shell Script Executor: Added Shell Script Executor support
- AppleScript Executor: Added AppleScript Executor support
Conch TTS Support πβ
- Added support for Conch TTS
Google LLM Provider Updates πβ
- Updated to latest gemini-2.0-flash-thinking-exp-01-21
- Native support for audio message types
Optimizations β‘β
- Improved TTS performance
- Optimized Python environment initialization
- Added parameter display in Workflows
- Other general improvements
EnConvo 2.0.3 Changelog (2024-12-20) π
Chat Window Enhancements π¬β
- New UI design for improved aesthetics
- Performance optimization with reduced CPU usage
- Expanded functionality: Integrated Voice Input, Screenshot, Web Search, Live Screen, Live Camera, File Attachment, and Realtime Call features directly within the Chat Window
YouTube Plugin Integration π₯β
- Chat With YouTube: Engage in conversations about YouTube video content
- Summarize YouTube: Generate concise summaries of YouTube videos
- YouTube Transcript Loader: Extract video subtitles , or save as SRT or TXT files
- YouTube Context Awareness: Smart recommendation of relevant commands based on currently open YouTube videos in the browser
- Source Code : https://github.com/Enconvo/YouTube
Code Runner Plugin π»β
- You can run shell, Node.js, Python, and AppleScript code
- Empowers users to extend EnConvo's capabilities using their preferred programming languages
- Source CodeοΌ https://github.com/Enconvo/Code-Runner