README.md

🎠 Motrix - macOS native chatGPT client

Add feature to backup all datas and restore back in preferences.
Add feature to duplicate message to another chat node
Add memo chat mode to only save something and not send ai server
Add tmp send message as group chat message by start new chat with format '@node1;node2;node5 newline translate text'
Add quick tmp group chat from left tree node's context menu
Add feature to lock context to row, let conversation always send from locked context row
Add double tap to edit timer tiny pop chat window to edtior mode
Add group chat parallel thread options in slide bar, it use to reduce parallel request to api cause server response issue
Add edit timer pop options direct inside pop chat window
Add shortcut to open any chat node with float window
Fixed quick window theme style not match
Fixed stream mode can't read data from open source project litellm with proxy mode
V1.9.2-Uni

Add double tap to edit timer tiny pop chat window to edtior mode
Add edit timer pop options direct inside pop chat window
Add shortcut to open any chat node with float window
Fixed quick window theme style not match
Fixed stream mode can't read data from open source project litellm with proxy mode
V1.9.1-Uni

Add support llm/gpt proxy deploy local (litellm) http://127.0.0.1:8000 as host to support most open source llm/gpt chat server
Add clone feature for chat tree node in right context menu
Add batch setup nodes model parameters feature (select multi-nodes from tree then use context menu -> batch setup)
Fixed system instructions description can't collapse issue
73.6 MB

Add shortcut to switch to next tree chat node (Use to combine with other system batch shortcut apps)
Add snapshot in chat content right context menu.
Schedule popup AI query result window can save to full result image to clipboard
Chat content row toolbar removed. Use right context menu as replacement
Fixed batch share chat content to image layout issue
Fixed crash on startup macOS 11
Download universal edtiton 73.5 MB

Added support new GPT3.5 & GPT4 16K model
Added support custom timer popup window default width (Use for long response daily ai tips)
Fixed stream mode some case display issue
Fixed the tiny popup timer window display bugs
Fixed macOS11.0 new node stuck issue

Add Shortcuts App support (macOS 13.0+) Query AI response then embed in to your self auto workflow. You can view how to use inside app's preferences
Motrix quick service now can select tiny window to show ai response content
Fixed Motrix service can't show on browser or other text app

Add daily timer to auto query ai chat node for popup inspirational or motivational tips, Or you can use it to daily auto popup an coding tips, Language practise. (Double tap on user question row)
Add daily timer can direct invoke from Motrix menubar's contextmenu
Add global shortcut CMD+SHIFT+F quick switch to global search favorites
Fix fix fix scroll issues
Fix pop template selector sometimes can't insert text
Stream mode performance improvements

Add float win mode, you can make an chat windown to an new tiny floating panel
Add support drag text to chat window any area to send message
Add show layout switch on main toolbar option & support custom layout restore icon
Add Motrix quick system service, When in other editor app select any text select Motrix quick will popup chat nodes selector and send selected text to the chat system
Add right mouse context menu on message for quick copy & fav & edit
Fix scrollview stable issue
Fix markdown newline bug

Add stream mode to get response tokens one by one
Add message embed edit by double tap the message row
Add appearance config in settings for some UI custom
Add 3 new themes with Light & Dark support
Add layout save & restore features (include current theme & lightDark & font size & auto options ...)
Add Typing animation option
Add os text service, call Motrix's api in all other system texteditor
Fixed & Tested Mortix is compatible with macOS 11.0, 12.0 devices
Bug fixed and Performance improvements

Support search in all conversations
Support batch select mode. select chat messages to share or export
Support show favorites in all conversations (Inside search menu)
Custom each node max tokens (You must know how these work, like context message,it can effect ai know how much of the the contextual)
Custom node default enter mode
Bug fixed and Performance improvements

Speech mode: you can use microphone to record text. App will translate to text and keep the audio track.
Voice speak content: Single click user or bot avatar.
Group chat mode: you can set parent node and add some child nodes (Must use system instructions to init ai role & it's role), then send message in parent node, All childs will give response in current conversation.
Auto clipboard: auto copy content to clipboard when response in
Auto speech: auto speak response
Auto mask: auto mask ai response. This is mainly for users who want to practice dictation of various languages.