v14.0.0

Ultra Plan Launch & Advanced AI Capabilities

Sree
SreeAuthor

This major release introduces our new Ultra plan, expands AI capabilities with advanced reasoning features, and brings free access to Provider 1 and Provider 6 for all users.

New Features

6
  • Ultra Plan Launched: Introduced new "Ultra" plan designed for high-volume users with enhanced features and capabilities.
  • Provider 1 & 6 Free Access: Made Provider 1 and Provider 6 completely free to use with original context limits and max tokens maintained for all users.
  • New AI Provider Integration: Added support for a new AI provider, giving access to another family of powerful and highly-capable models.
  • Expanded Reasoning Capabilities: Advanced reasoning feature now supported by a wider range of models across integrated providers, allowing for more complex problem-solving.
  • Verbosity Parameter Support: Added verbosity parameter to select models for controlling response detail level.
  • Free Tier Experimental Models: Added cutting-edge experimental models for Free tier users:

Improvements

3
  • Improved Playground Performance: Enhanced timeout handling and HTTP client management within the Playground for a more responsive and reliable experience.
  • Provider 7 Enhancements: All Claude models now available in Provider 7 (Ultra plan exclusive) with reasoning effort support: Low (8K), Medium (16K), High (32K tokens).
  • Enhanced OpenAI Models: Added "reasoning effort" parameter support and timeout extended from 60s to 600s for better performance.

Bug Fixes

3
  • Enhanced Tool Handling: Resolved issues to improve implementation and reliability of tool-handling features for core providers.
  • Qwen Models Fixed: Resolved stability issues with Qwen models.
  • Gemini 2.5 Pro Fixed: Addressed issues affecting Gemini 2.5 Pro model.

Important Notes

1
  • Provider 1 Rate Limiting: Some models may experience temporary downtime due to rate limiting. If a model is unavailable, try another model from the same company in Provider 1.