Feature Requests & Development Status
See what we're working on and what's coming next for MxChat
🚧 Currently In Development
Features we're actively working on right now
Claude PDF API Integration
Add Claude\'s native PDF processing capabilities to enable advanced document analysis including visual content understanding, chart interpretation, and structured data extraction. This will provide users with AI-powered document insights beyond basic text extraction.
Enhanced PDF Processing with PDF.CO
Add PDF.CO API as an advanced processing option to handle complex PDF formats and improve data extraction accuracy for knowledge base ingestion. This will provide users with a more robust alternative for challenging documents, scanned files, and varied layouts.
🔍 Under Review
Features being planned and reviewed for development
OpenRouter API Integration
Add OpenRouter API support to provide users with access to multiple AI models (GPT-4, Claude, Gemini, etc.) through a single unified interface. This will include model selection dropdown, embedding model support for enhanced semantic search, and competitive pricing options, giving users flexibility to choose the best model for their specific use case while maintaining familiar OpenAI-compatible functionality.Retry
Add GPT-5 to Theme Customizer
Integrate GPT-5 models into the theme customizer add-on for enhanced customization options.
🚀 Released Features
Features that have been released and are live in MxChat
Version 2.4.4 3 features
Pinecone UI Database Enhancement
Currently Pinecone database UI is cached with an update every 60 seconds. This is causing confusion for users when entering or deleting records as they do not update right away. Improve cache busting mechanism to always show current information.
Similarity Threshold Fix
On a fresh install the backend shows the similarity threshold at a default of 35, but the frontend debugger shows a threshold of 75. These should be aligned on initial load to 35.
Fix Missing Characters in Response
Implement a fix for characters being lost when streaming is enabled. This causes words to become jumbled together or characters missing from words.
Version 2.4.0 3 features
Essential Elements Error
When you have the \\\"Require email to chat\\\" option turned off the dev console shows \\\"Essential elements for email handling are missing\\\". It should check first if the option is enabled/disabled before checking if the elements are missing.
Chat Persistence Error
When chat persistence is enabled and the user has no chat history it will show \\\"error chat history not found\\\".
Slack Live Agent Switching
Bug Fix: When switching to live agent and back to chatbot, the agent indicator does not show its back on the chatbot until the page is refreshed. It should show switch as the status switches.
Version 2.3.9 7 features
Slack Channel Naming
Update Slack channel naming logic to prioritize using the client\\\'s name, falling back to email, and then chat ID if unavailable.
DeepSeek Streaming Responses
Add streaming response support for DeepSeek models.
Direct Content No Source
Allow users to submit multiple direct content entries without a source URL.
Fix Auto Sync Drafts
Ensure posts moved from posted to draft are removed from the knowledge database, addressing the current retention issue.
Role-Based Chatbot Info Access
Implement a feature to restrict chatbot information access based on user roles for enhanced security and relevance.
Add Chat Hide Option
Implement a feature in the admin panels post edit page to hide the floating chatbot on specific pages.
Product Card Action
Add support for variable products and shorten AI response text in product card actions.
Version 2.3.8 3 features
Fix Extra Characters in Knowledgebase
Prevent extra backslashes from appearing when editing content in the knowledgebase.
Require Email to Chat Form
Add optional name field to the require email to chat form.
Slack Channel Names
Review and improve Slack channel naming conventions to make chat tickets more distinguishable.
