May 9th, 2025

This release focuses on enhancing AI configuration options and improving the user interface for better usability.

GenStudio

Adjust AI Function Call Aggressiveness: Users can now control how aggressively the AI utilizes function calls, tailoring interactions to their preferences.
Configure Claude 3.7 Sonnet Thinking Parameter: Access and adjust the thinking parameters of Claude 3.7 Sonnet for customized AI responses.
Behavior Settings in Right Sidebar: Easily access and modify AI behavior settings directly from the right sidebar.

Collapsible Walkthroughs and Thinking Process: Walkthroughs and the AI thinking process are now collapsible for a cleaner interface.
Model Filtering: Users can now filter models when selecting, making it easier to find the desired model.
Improved Behavior Settings Panel: Enhanced the right panel for behavior settings for better usability.
Adjustable Panel Sizes: Users can adjust the minimum sizes of the left and right panels for a customized layout.

Fixed UI Spacing Issue: Resolved a spacing issue after the "Apply" button.
Tool Name Length Limitation: Ensured tool names are limited to 63 characters to support Gemini models.
Support for Duplicate Tool Names: Allowed duplicate tool names for function calling by prefixing with server names.

Adjust AI Function Call Aggressiveness: Users can now control how aggressively the AI utilizes function calls, tailoring interactions to their preferences.
Configure Claude 3.7 Sonnet Thinking Parameter: Access and adjust the thinking parameters of Claude 3.7 Sonnet for customized AI responses.
Model Switch Recommendations: The system now advises users to switch models when appropriate.

Re-enabled Function Calling for Groq Model: Restored function calling capabilities for the Groq model.
Improved System Prompt for Moderate Behavior: Enhanced system prompts to encourage more moderate AI responses.
Conversation List Enhancements: Improved the conversation list for better performance and usability.

Fixed Groq Model Failure with Tool Calls: Resolved an issue where the Groq model would fail if the message history contained a tool call.
Gemini Image Generation Improvements: Addressed issues with image generation in the Gemini model.
Resolved Message Content Error: Fixed an error where message content could cause failure.
Removed Redundant Conversation Attributes: Cleaned up redundant attributes in the conversation model to improve performance.

Last updated 1 year ago