Grok AI Toolkit v1.1.0

Update Vision Mode

[1.1.0] - 2024-12-28

Added

• Support for image detail parameter in vision mode
• File size validation (10MB limit) for image uploads
• Proper message structure formatting for xAI Vision API
• Documentation for all supported models and their capabilities

Changed

• Updated FileHandler to handle image detail parameter:
  • process_files() now accepts detail parameter
  • process_file() passes detail to appropriate processing functions
  • handle_upload_command() supports detail parameter
• Modified GrokBase class:
  • handle_user_input() now accepts image_detail parameter
• Updated Vision class to properly handle image parameters
• Improved message content structure to match xAI API specifications
• Moved validation logic to InputValidator class
• Updated model documentation with latest context windows:
  • grok-2-1212: 131,072 tokens
  • grok-2-vision-1212: 8,192 tokens
  • grok-vision-beta: 8,192 tokens
  • grok-beta: 131,072 tokens

Fixed

• Fixed argument mismatch in process_files()
• Corrected message structure for text-only responses in vision mode
• Standardized image URL and base64 message formatting
• Updated file size limit from 20MB to 10MB to match xAI specifications

Technical Details

• Image detail parameter defaults to "high"
• File size limit enforced at 10MiB (10 * 1024 * 1024 bytes)
• Message content now properly structured as array of type-specific objects
• Added model aliases documentation for grok-2 and grok-2-latest

Full Changelog: v1.0.0...v1.1.0

Grok AI Toolkit v1.0.1

• Change repo name

Grok AI Toolkit v1.0.0

Initial release.