AI-Starter 2.4.3
We are pleased to announce the release of the latest version of AI-Starter, version 2.4.3. This update focuses on improving usability and functionality, including a refreshed user interface and the addition of PDF file input capability, which many users have requested.
Main Update Contents
- New logo and UI
- Support for GPT-4o mini
- Enhanced file input functionality
- Other minor improvements and optimizations
1. New Logo and UI
We have implemented the new logo and UI that we previously announced
- Applied the AI-Starter logo
- Improved message input field
- Added a help button
- Enhanced conversation display
These changes are the first step in a long-term improvement plan aimed at creating a more intuitive and user-friendly UI. We plan to continue releasing new UI elements in stages.
2. Support for GPT-4o mini
We have added support for GPT-4o mini, a small model gaining attention for its cost-effectiveness. It is available through both the OpenAI API and Azure OpenAI API.
For more details, please refer to the following links:
3. Enhanced File Input Functionality
To improve the convenience of file input, we have made the following improvements:
- Added a file upload button to the message input field
- Traditional copy & paste and drag & drop attachments are still available
- Support for image input with Gemini 1.5 series assistants
- Support for document file and spreadsheet file input with Claude 3 series assistants (Haiku, Sonnet, Opus)
- Utilizing the Document chat feature in Amazon Bedrock's Converse API
- Claude 3.5 Sonnet does not support the Document chat feature, so this functionality is not available
- For more details, see File Attachment - Supported File Formats.
Reference information:
Conclusion
AI-Starter values feedback from our users and continues to make improvements. Please try out the new features and share your opinions and requests with us. Your feedback contributes to the further evolution of AI-Starter.
We look forward to bringing you more updates in the future.