v0.14.0

🚀 Release Notes - May 6, 2026

✨ New Features

• Fireworks AI Provider: Added support for Fireworks.ai as a new LLM provider, expanding model availability options

🔧 Improvements

• LLM Model Registry: Refactored model list to only register models that have LLM capabilities, improving initialization efficiency
• Token Management: Reduced default max_token limit to 80k to trigger context shrinking sooner and prevent overflow issues
• Usage Chunk Processing: Moved usage chunk detection earlier in the pipeline for better tracking and monitoring

🐛 Bug Fixes

• Console Performance: Removed unnecessary refresh operations in the console, improving UI responsiveness
• NVIDIA Model Capability: Fixed correctness reporting for NVIDIA model capabilities

🔄 Model Updates

• Qwen Model: Replaced partner model with open-source Qwen variant for better accessibility and licensing