Skip to main content

🧚 Some updates have dropped

Released on May 3, 2026

This release introduces regional endpoints with full inference control — giving you total control over where your AI workloads run, from individual agents to workspace-wide enforcement! 🚀

🌍 Regional Endpoints — Your Compute, Your Region​

Total Control Over Where Your AI Workloads Run​

At ToothFairyAI, we've always championed regional sovereignty — your data, your rules, your region. Historically, when capacity ran low in your preferred region, we had to route some requests to nearby regions to maintain availability. Today, that trade-off is gone. We now offer regional endpoints that give you granular control over exactly where your inference happens.

Why This Matters: Imagine your data is sensitive, your compliance requirements are strict, or your contracts mandate that all AI processing happens within Australian borders. With regional endpoints, you decide precisely where your compute lives — no compromises.

Two Routing Modes:

ModeBehaviourBest For
Global (default)Prioritises your preferred region but routes to Europe and/or US if model availability in your region drops below 50%Most use cases — prioritises your region while ensuring availability
RegionalRequests stay in your selected region NO matter the degradation level. If capacity is unavailable, the request is queued until capacity frees upMaximum data residency, strict compliance, contractual obligations

Real-Time Model Availability: You can monitor model availability across regions in real-time directly in the ToothFairyAI app. Navigate to model selection in any agent to see current availability percentages for each region — making it easy to decide whether to use Global or Regional mode based on current conditions.

How to Set It:

  1. Navigate to Settings > Regional
  2. Select your preferred region (e.g., Australia East)
  3. Choose your routing mode (Global or Regional)
  4. Save — all future requests honour your preference

Availability: Regional endpoints are available for all subscriptions. The default routing mode is Global.

Workspace-Level Enforcement: Admins of Business and Enterprise workspaces can enforce regional inference across all agents in the workspace — preventing data leakage from misconfigurations. This ensures that even if an individual agent is set to a different routing mode, all inference stays within your configured region.

Why This Matters for Admins:

  • Prevent Misconfiguration Leaks - A single agent misconfigured to "Capacity First" won't accidentally route sensitive data outside your region
  • Compliance Assurance - Enforce regional sovereignty at the workspace level without trusting each team member to configure correctly
  • Audit Simplicity - One workspace-wide setting guarantees compliance rather than auditing individual agent configurations

How to Enforce:

  1. Navigate to Settings > Workspace > Regional
  2. Enable Enforce Regional Inference
  3. Select the required region for all agents
  4. Save — all agents in the workspace are now forced to regional inference

📖 Documentation: Regional Endpoints

🎯 Agentic Tooling — Smarter Web Scraping & Error Recovery​

Direct URL Scraping & Intelligent Error Handling​

We've enhanced Agentic Tooling with two powerful improvements that make your agents more effective at gathering information and recovering from execution issues.

Direct URL Scraping: Agents can now scrape specific URLs directly — bypassing the need for web search when you already know where the information lives. Simply provide a URL and the agent fetches, parses, and analyses the content with full context awareness.

What This Enables:

  • Skip the Search - When you have a direct link (documentation, article, product page), the agent goes straight there
  • Full Page Analysis - Extract structured information from complex pages including tables, lists, and formatted content
  • Authenticated Content - Access pages behind authentication when credentials are configured
  • Batch Scraping - Agents can chain multiple URLs together for comprehensive research without search overhead

Error Recovery — Smarter Suggestions: When an agent encounters an execution error, it no longer simply retries with the same approach. Instead, it now analyses the failure, suggests alternative paths, and proposes solutions — keeping your workflows moving forward.

How Error Recovery Works:

  • Root Cause Analysis - The agent identifies why the tool call failed (timeout, rate limit, invalid input, permissions, etc.)
  • Alternative Paths - Suggests different approaches that might succeed where the first attempt failed
  • Smart Retry - Offers adjusted parameters for retry rather than blindly repeating the same request
  • Escalation Guidance - When the agent can't recover, it clearly explains what went wrong and what information you need to provide

Example: Instead of: "Tool call failed. Retrying..." Your agent now says: "The web search timed out due to high traffic. I can try: (1) Retry in 30 seconds, (2) Use a different search engine, or (3) Scrape the specific URL directly if you have one. Which would you prefer?"

📖 Documentation: Agentic Tooling 2.0

🖥️ TF Code — Deeper ToothFairyAI Integration​

Faster, Smarter, More Integrated​

TF Code is now more deeply integrated with ToothFairyAI than ever before — faster loading, pre-configured for your workspace, and with improved MCP handling for seamless automation.

What's Improved:

  • Faster Loading - TF Code now starts up significantly quicker, so you can get to work without the wait. We've optimised the initialisation process to get you coding in seconds, not minutes.

  • TF Engineer — Pre-Configured for Your Workspace - A dedicated TF Engineer agent is now pre-configured out of the box, designed specifically to complete tasks in your ToothFairyAI workspace. It understands your workspace context, knows your agents, documents, and tools — ready to help you build, automate, and manage without additional setup.

  • Better TF MCP Handling - The ToothFairyAI MCP integration has been refined for more reliable connections and smoother tool execution. Authentication is faster, connection drops are handled gracefully, and tool responses are parsed more efficiently.

What You Can Do:

  • Instant Automation - Start automating ToothFairyAI tasks immediately with the pre-configured TF Engineer
  • Workspace-Aware - The agent knows your workspace structure, agents, and available tools
  • Reliable Connections - MCP connections are more stable with automatic reconnection and better error handling

📖 Documentation: TF Code

📱 ToothFairyAI Mobile — Coming Soon to iOS & Android​

Your AI Agents, Now in Your Pocket​

We're excited to announce that ToothFairyAI is landing on your mobile devices — both iOS and Android! The app brings the full power of ToothFairyAI to your phone or tablet, so you can stay productive on the go.

What's Included:

  • Full Agent Interaction - Chat with your agents directly from your mobile device, just like on desktop
  • Dispatch Work - Send tasks to your agents and let them work asynchronously while you move on with your day
  • Real-Time Notifications - Get notified the moment your agent completes a job, so you're always in the loop without having to check manually

Launch Timeline: We're currently waiting for Apple's validation before publishing to the App Store. Rest assured, we'll be coordinating the iOS and Google Play releases to make sure you can access ToothFairyAI on both platforms at the same time.

Keep an eye on the app stores in the coming days — your AI agents are about to become even more accessible!

📖 Documentation: Mobile App

🛠️ Improvements & Fixes​

European Login Fix​

Fixed an issue that was preventing European users from logging into the ToothFairyAI app. Authentication flows have been updated to properly handle European region configurations, ensuring seamless access for all users regardless of their location.

📋 Summary​

This update includes:

  • Regional Endpoints - Full control over inference location with Global and Regional modes
  • Workspace-Level Enforcement - Business/Enterprise admins can force regional inference on all agents, preventing data leakage from misconfigurations
  • Agentic Tooling Enhancements - Direct URL scraping and intelligent error recovery with alternative path suggestions
  • TF Code Integration - Faster loading, pre-configured TF Engineer for your workspace, and improved TF MCP handling
  • Mobile App - ToothFairyAI launching on iOS and Android with agent interaction, work dispatch, and real-time notifications (pending Apple approval)
  • European Login Fix - Resolved authentication issue preventing European users from logging into the app

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! 🧚‍♀️✨

v0.679.0