May 16, 2024

Introducing Enhanced AI Capabilities and Flexible Training Options

๐Ÿš€ New Features
  • Latest OpenAI GPT-4o Model: We have integrated the latest OpenAI GPT-4o model, the best in its class, now available to all users with access to GPT-4 series. This model offers a variety of context windows to suit diverse needs.
๐Ÿ’ก Improvements
  • Agent Retrain Approval: Previously, the AI Supervisor intents were automatically retrained with each update to the Agent name, description, or prompt. Based on user feedback, we've introduced a notification system that asks if bot makers would like to regenerate AI Supervisor intents on save, providing more control and flexibility.
To learn more about fine tuning intents, please visit:
May 8, 2024

Seamless Chatbot Management and Expanded Notification Options

๐Ÿš€ New Features
  • Transfer Chatbot Ownership: Users with the Pro plan can now transfer chatbot ownership to other users, provided that their subscription support's all of the chatbot's features.
๐Ÿ’ก Improvements
  • Email Notifications: You can now add multiple email addresses to send leads / variables data to multiple recipients.
May 2, 2024

Email Notifications and Real-Time Assistance

๐Ÿš€ New Features
  • Email Notifications: Take your Sales / Customer Support workflow to the next level with the introduction of email notifications. GPT-trainer chatbots can now efficiently send email notifications containing your newly collected leads data. Find email notification settings in the Settings / Appearance section.
    • Daily Chat History: Stay updated with a daily email summary of the day's chat session history, ensuring you never miss important conversations.
    • Daily Collected Data History: Receive a daily email summary of all collected user data (or any variables you defined) across all chat sessions.
    • On-Demand Lead Forwarding: Get immediate access to newly collected leads' information or selected variables via email, enhancing your lead conversion.
  • Human Escalation Notifications: Our Human Escalation Agent now sends notifications through your browser when users request to chat with a real person. Whether you're actively using our app or taking a break, you'll receive notifications in any tab within your browser when someone requests human agent assistance. To set this up, refer to our dedicated guide here.
  • Unread Messages: Never overlook important messages again. We've added a badge to the Inbox sidebar tab to show the number of unread messages.
๐Ÿ’ก Improvements
  • Alerts UI: Improved UI for popup alerts.
April 29, 2024

Expanding Functionality and Refining User Experience

๐Ÿ’ก Improvements
  • Free Advanced Scraper: We're excited to announce that our Advanced Scraper is now available for free to all users! All website scraping will automatically be done by the Advanced Scraper. No action is needed on your part.
  • As our platform becomes more popular, managing system resources effectively is crucial to maintaining performance and fairness for all users. That's why we have introduced some updates to scraper rate limits:
    • Auto Re-train Frequency: To optimize resource allocation, daily re-training options have been removed. Available frequencies are now weekly, biweekly, and monthly.
    • URLs Per Day: We have implemented a daily limit of 2000 URLs per user. If this limit is reached, an alert status will be displayed, and additional URLs can be added the next day.
    • Auto Re-train: If you have 3000 URLs stored, then when retrain happens, the first 2000 URLs will be retrained on the day marked for retrain, and the remaining 1000 will be done the following day.
  • Multiline Initial Messages: Enhancing message flexibility, users can now input multiline initial messages. To create separate messages, simply include two line breaks, otherwise, consecutive lines will be treated as part of a single message.
  • UI Updates: We've rolled out several UI enhancements aimed at improving user interaction and aesthetics:
    • Redesigning Add Sources Button: We have transitioned from a dialog to a dedicated page, enhancing usability and allowing for a larger interface. This change primarily enhances user experience by making it easier to navigate and interact with the feature.
    • Chat Citations: Minor improvements have been made to colors and sizing for better readability.
    • Widget Animations: Experience smoother transitions when opening and closing widgets, adding a touch of elegance to your interactions.
    • Integrations Clarity: We've made it easier to understand what each integration entails, improving user experience.
๐Ÿž Bug Fixes
  • Link Sharing: We've resolved an issue in link sharing where disabling authentication was incorrectly defaulting, hindering the sharing process.
  • Chat Input: Fixed a bug where the chat input was automatically removing text following AI-generated responses, ensuring your full message remains intact.
April 23, 2024

Expanding Functionality and Refining User Experience

๐Ÿ’ก Improvements
  • Agent Redesign: Introducing a new agent type - Fixed Response. This agent can defend against malicious input, prompt injection attempts, or queries unrelated to your main chatbot functions. It does not require any Message Credits to run since it outputs a fixed message. Minor settings updates have been made for Background agents for clarity.
  • Advanced Scraper Enhancements: Advanced Scraper now supports sitemap extraction. Weโ€™ve also improved error handling to provide a smoother experience.
  • Sitemap Size Increase: We have increased the limit for sitemap extraction from 500 to 1000 links, enhancing our ability to handle larger websites.
๐Ÿž Bug Fixes
  • Trial Date Correction: Fixed an issue where incorrect trial dates were being displayed.
  • Menu Sidebar Scroll Fix: Resolved a problem preventing the sidebar from scrolling on small devices.
April 12, 2024

Enhancing Capabilities and Revamping Subscriptions

๐Ÿš€ New Features
  • Google Search Integration: AI Agents now have a new knowledge source option. Enable this option for your Agent to run a Google search for additional context as it generates a response. Please note that not all results on Google are true or accurate, as it pulls information from the internet at large.
  • Advanced Scraper: Use the โ€œAdvanced Scraperโ€ by selecting the associated checkbox when adding a website URL or assign it to process existing sources during automatic retraining. Advanced scraper uses a server-side browser rendering technology in order to capture some dynamic content.
  • While this makes scraping more reliable in general, we still cannot guarantee that it will work for all websites.
  • Feature Credits: support powerful features like Google Search and Advanced Scraper, we are introducing Feature Credits (FCs). Each account will receive a one-time gift of 10,000 FCs to test these features. If you find them helpful, you can purchase monthly add-ons for $9.99 per 100,000 recurring FCs.
  • Data Collection Webhook: Connect to your favorite CRM with our new webhook. Select the Variables you want to send to your API endpoint. Once the AI collects all the needed information, it will be sent automatically. Configure this webhook in your chatbotโ€™s Settings / Appearance tab.
  • SOC II Compliance: GPT-trainer is officially SOC II: Type 1 certified! We are in the process of going through audits right now for ISO 27001:2022 and working towards full GDPR compliance. We estimate to achieve both by the end of May, 2024.
๐Ÿ’ก Improvements
  • Auto-Retrain Customization: Previously, you could only retrain all URLs at once. Now, select individual URLs for retraining and set the frequency of retraining to daily, weekly, or monthly.
  • Subscription Overhaul: Following recent communications, weโ€™ve significantly redesigned our public-facing subscriptions. Free Plan now becomes a 14-day trial with unlimited access to most features and a one-time 2,000 MC assignment. After the trial expires, Free Plan users will need to purchase a paid subscription to continue usage. If you are a legacy subscription holder and would like to switch to one of our new subscriptions immediately, please email us at Questions about the transition can be directed to our support team at
๐Ÿž Bug Fixes
  • Duplicate Sources Error Handling: Improved error messaging for attempts to upload duplicated sources.
April 5, 2024

Elevating Chatbot Performance and Expanding Customization Options

๐Ÿš€ New Features
  • Human Escalation: Enhance user interaction with real-time human support. Add the Human Escalation Agent and activate it for your embedded chatbots using iframe or script. When a user requests to talk with a human, the chat will be marked as Support Requested, alerting human agents on standby to take over the conversation right away. This feature starts from the Standard tier.
  • Use Your Own OpenAI Key: Add your own OpenAI API key in the Account menu section. Enjoy unlimited chatbot usage by providing your own OpenAI API key for interactions.
  • Per-Chatbot Message Credits Limit: Set a monthly limit on message credits under Security Settings to control your chatbot's AI costs. Once the limit is reached, the chatbot will automatically be disabled and respond with a pre-canned message.
  • White-label Entire Platform: Apply for a free commercial license from us and launch your own chatbot platform! Invite customers, earn revenue, and become a part of our thriving community. Customize the platform to your liking and take your business to new heights. For partnership inquiries, reach out to us at
๐Ÿ’ก Improvements
  • UI Updates: Implemented UI updates to enhance compatibility with the white-label solution.
  • Redesigned Inbox UI: Updated the Inbox interface to deliver a better chat session monitoring experience.
  • Initial-Agent and Required Variables Settings: With our AI Supervisor Overrides allowing custom rules for agent switching, weโ€™ve removed the need to set an initial agent. Every setup still requires at least one User Facing agent, but the "initial" agent will now be decided automatically by AI or via your provided rules. Please review our guide for setting up AI Supervisor Overrides on our documentation page for more details: Supervisor Overrides
๐Ÿž Bug Fixes
  • View Content: Fixed an issue where users couldn't see the correct source content after retraining.
  • Custom Domain: Added an appropriate error message for custom domain duplication.
March 12, 2024

Introducing Advanced AI Supervision and Enhanced User Experience

๐Ÿš€ New Features
  • Agent Intents: Leverage direct use of intents by the AI Supervisor for improved relevance calculation based on the chatbot's latest input. Editable intents allow for fine-tuned task delegation.
  • Agent Bias/Stickiness: modifiers for default relevance, with Bias for general preference/aversion and Stickiness for added switching cost, enhancing the AI Supervisor's decision-making.
  • Download QA Sources: Ability to download QA sources as JSON files for easier data management and integration.
๐Ÿ’ก Improvements
  • Revamped Documentation: We've listened to your feedback and have completely overhauled our documentation. Start building your chatbot from the ground up with our comprehensive guides, covering everything from initial setup to advanced customization. Explore our new resources in GPT-trainer Documentation. Though we are still adding more guides for everyone to make sure your journey on using our tool will be easier and joyful.
๐Ÿž Bug Fixes
  • Conversation Appearance: Fixed an issue causing old conversations to appear at the top. Now, only conversations with new messages will be prioritized.
March 8, 2024

Enhance Your AI Chatbot Capabilities and Knowledge

๐Ÿš€ New Features
  • Agent Rule Overrides: Improve control over AI Supervisor with Supervisor Overrides. Apply conditional logic to precisely define when an agent should take over a conversation.
  • Metadata for Functions: New toggle for including default metadata as function payload. When activated, the Agent will include information such as session uuid as payload during function calls.
February 29, 2024

Customizable Pre-Canned Responses

๐Ÿš€ New Features
  • Pre-Canned Responses for Suggested Prompts: Users can now provide their own standard responses for suggested prompts of the โ€œUser Messageโ€ type. Automatic delivery of these responses will not incur any Message Credit cost. You can leave this field blank, in which the AI will generate its own response using Message Credits as usual.