Skip to main content

What's New in Version 0.672.0

Released on March 3, 2026

This release brings expanded voice capabilities with enhanced multilingual support, powerful versioning for agents and tools, improved orchestrator reliability, and significant API enhancements! 🚀

🎙️ Voice Agents: Expanded Language Support​

Native Multilingual Capabilities​

Voice agents now speak many more than 15 languages natively, offering a wide range of accents to create truly global conversational experiences. And yes—you guessed it—Australian voices are now available too!

Key Features:

  • 15+ Native Languages - Comprehensive language coverage for global deployments
  • Wide Range of Accents - Regional accents that sound authentic and natural
  • Australian and European Voice Support - Aussie and European languages and accents now available for localised experiences

This expansion enables you to deploy voice agents worldwide with native-sounding voices that resonate with local audiences.

🎯 Enhanced Sorcerer Models for Customer Service​

Improved Steerability & Speed​

We've enhanced the customer service steerability from Sorcerer models while simultaneously ensuring increased speed when generating answers. This dual improvement delivers faster, more controllable responses.

What's New:

  • Enhanced Steerability - Fine-grained control over response tone, style, and behaviour
  • Increased Generation Speed - Faster response times without sacrificing quality
  • Non-Agentic Optimisation - Particularly effective for direct customer service use cases where you simply want to create a customer service agent or concierge
  • Voice Agent Base Model - Improved performance when using Sorcerer as the base model for Voice agents

Use Cases:

  • Customer Service Concierge - Direct Q&A without complex agentic workflows
  • Voice Agent Foundation - Ideal base model for voice-enabled customer interactions
  • Support Desk Automation - Fast, steerable responses for helpdesk scenarios

📋 Agents, Tools & Hooks Versioning​

Full Production Release​

Agents/Tools/Hooks versioning is now fully live and production ready, giving you complete visibility and control over how your AI resources evolve over time.

Key Capabilities:

  • Complete Version History - Easily review how an agent has been updated over time
  • Change Inspection - See what changes were made at which point in time
  • Real-time Diffs - View differences between versions in real-time
  • One-Click Rollback - Instantly restore previous configurations with a single click
  • Continuous Adjustment - Keep fine-tuning settings as needed while maintaining version control

Availability: This feature is available exclusively for Business and Enterprise subscriptions.

This enterprise-grade versioning system brings software development best practices to your AI resource management.

🔌 MCP Updates for Coding Agents​

Enhanced SDK & API Integration​

We've updated our MCP (Model Context Protocol) implementation to better enable coding agents to use our SDK and API accurately, as more customers leverage our SDK to create robust production-ready agents.

Improvements:

  • SDK-Specific Optimisations - Better context and examples for ToothFairyAI SDK usage
  • API Accuracy - Improved agent understanding of API endpoints and parameters
  • Code Generation - More accurate code generation for agent integration patterns
  • Production-Ready Patterns - Templates and examples for robust agent implementations

This enhancement helps developers build more reliable agents that correctly leverage our platform capabilities.

🛠️ Orchestrator Parallel Task Fix​

Resolved Cascading Execution Issues​

We've fixed a critical issue where parallel tasks running as part of an orchestrator agent's plan execution would not complete, causing a cascading effect on overall execution.

This fix ensures your orchestrator agents can handle complex parallel workflows with confidence.

🔧 API Tools Robustness​

Dynamic Data Conversion​

We've improved API tools robustness by enabling agents to dynamically convert base64 and raw binary data into proper text for analysis and answer generation.

New Capabilities:

  • Base64 Decoding - Automatic conversion of base64-encoded data to readable text
  • Binary Data Processing - Raw binary data can now be analysed and interpreted
  • Files Analysis - For example, a PDF served over a GET API can now be analysed directly by agents
  • Document Understanding - Extract and process content from binary file formats

Example Use Case: An agent can now call an API that returns a PDF document, automatically convert the binary data, and analyse its contents to generate answers—all without manual intervention.

📋 Summary​

This update includes:

  • Voice Agent Languages - 15+ native languages with wide accent range including Australian voices
  • Sorcerer Enhancements - Improved customer service steerability with faster response generation
  • Versioning System - Full production release of Agents/Tools/Hooks versioning for Business/Enterprise
  • Orchestrator Fix - Resolved parallel task completion issues with dynamic agent generation
  • MCP Updates - Enhanced SDK and API integration for coding agents
  • API Tools - Robust binary and base64 data conversion for document analysis

Thank you for using ToothFairy! We're constantly working to improve your experience. If you have any feedback or encounter any issues, please don't hesitate to reach out to our support team at support@toothfairyai.com.

Keep building magic with AI! 🧙‍♀️✨

v0.672.0