Release Note
V3.0(Apr 4th, 2025)

  • Design Overhaul

    ModelHub has been redesigned to align with the Motif’s new branding and look & feel. The design system has also been refined to provide a more consistent user experience.

V2.12(Mar 21st, 2025)

  • Auto-Translation for Prompts in Art Assistant Service

    Previously, if the Prompt Rewriter feature was turned off in the Art Assistant service, prompts had to be written in English for image generation to work properly. Now, prompts written in Korean are automatically translated into English, allowing smooth image generation without requiring manual translation.

  • Broken Image Filtering Feature

    Occasionally, broken images may be generated during the image creation process. With this update, such images are automatically identified and hidden, so users can view only the successfully generated results.

  • LaTeX Formula Visualization

    When you enter symbols using LaTeX syntax, they are now automatically rendered as visual mathematical formulas.

V2.11(Mar 7th, 2025)

  • Automatic Conversation Title Generation

    When starting a new conversation in the chatbot or medical consultation service, a title is automatically generated based on the first response and saved in the conversation history panel.

  • Saving Search-Based Response History

    Previously, when receiving search-based responses, the sources and related information were not saved in the conversation history. This issue has been fixed, allowing users to review source details and related content from past conversations.

  • Allowing Input During Response Generation

    Previously, input was restricted while a response was being generated. Now, users can enter prompts even while a response is still being processed.

V2.10(Feb 21st, 2025)

  • Service Stabilization Improvements

    In this version, we focused on enhancing service stability and performance by improving the backend architecture instead of adding new features. The internal system has been optimized to provide a smoother user experience, and we will continue making improvements in future updates.

V2.9(Feb 7th, 2025)

  • Prompt Rewriter Feature

    The Art Assistant service now supports the Prompt Rewriter feature. This feature enhances user-written prompts to generate higher-quality results.

  • Image History Storage Feature

    Images generated in the Art Assistant service are now saved, and previously created images can also be accessed from the left history panel. Currently, up to 10 images can be stored, but this policy is subject to change without prior notice.

  • Conversation History Storage Feature

    Conversations in the Conversational and Medical Consultation services are now saved, allowing users to continue previous discussions. Currently, up to 20 conversations can be stored, but this policy is subject to change without prior notice.

V2.8(Jan 17th, 2025)

  • T2I Service Launch

    A new T2I service has been added to the Labs section. You can easily create your desired images by entering prompts in English.

  • Search-Based Response Generation (Conversational AI Service)

    Similar to the Medical Consultation Service, the Conversational AI Service also provides search-based responses for your queries.

  • Usage Limit Policy for Search and Image Generation

    The search and image generation features are available only to registered and logged-in users. Both features have daily usage limits, which are reset at midnight KST (Korean Standard Time) each day. Please note that the usage limit policy is subject to change without prior notice.

V2.7(Dec 27th, 2024)

  • Search-Based Response Generation (Only for Medical Consultation Service)

    In the Medical Consultation Service, we provide search-based responses along with the referenced pages to enhance user understanding. Currently, the number of searches is limited, and once the limit is reached, answers will be generated without search results, without additional notifications. In the next V2.8 version, the same feature will be added to the Conversational AI Service, and the search usage policy will be further clarified.

  • Account Deletion Function

    You can delete your account registered with ModelHub through Google integration, effectively disconnecting it from your Google account.

  • Terms of Use and Policy Page

    We have enhanced transparency and user trust in the service by specifying essential information, rights, and policies required for users to utilize ModelHub, including the Privacy Policy.

V2.6(Dec 13th, 2024)

  • Release Notes Page and Announcement Modal

    A Release Notes page has been introduced, allowing users to view not only the latest updates but also the history of each version. This enhancement helps users better understand Model Hub and provides a clear overview of the progress we’ve made.

  • Improved Scrolling UX

    We’ve refined the auto-scroll behavior to ensure a smoother experience for users, especially when navigating long responses or scrolling during response generation.

  • Bug Fixes

    Resolved a bug related to editing sent questions where the history of the original question was not properly retained or displayed.

V2.5(Nov 29th, 2024)

  • Edit Sent Questions

    You can now edit sent questions and receive a new response. The original question and answer are preserved, allowing you to continue the conversation seamlessly.

  • Improved Medical Consultation Model

    We have enhanced the model's performance to provide more structured and detailed responses, improving the overall consultation experience.

  • Code Copy Feature

    For answers containing code, a 'Copy Code' feature has been added, making it quick and easy to copy the provided code snippets.

V2.4(Nov 15th, 2024)

  • Sign-Up and Login Features

    You can now sign up and log in using Google integration. Additionally, you can update your language and theme preferences through the profile settings menu.

  • Stop Response Generation

    If an undesired response is being generated, you can stop it immediately without waiting for the full response to complete, allowing you to resume the conversation right away.

  • Improved Text Hierarchy for Long Responses

    When generating long responses with bullet points or numbering, the text hierarchy has been improved for better readability and organization.

V2.3(Nov 1st, 2024)

  • Improved Dropdown Design for Model Selection

    The dropdown menu for selecting models has been redesigned for better intuitiveness. Simple tooltips have also been added to help users easily understand each service.

  • Positive/Negative Feedback Feature

    A new feature has been introduced, allowing users to provide positive or negative feedback on the responses they receive.

  • Enhanced User Experience for Long Responses

    When generating long responses, the user experience has been improved to prevent the screen from constantly shifting and to allow users to comfortably read the content while the response is still being generated.

V2.2(Oct 18th, 2024)

  • Launch of Medical Consultation Service

    The medical consultation service has been launched in the Labs section.

  • Success/Failure Toast Messages for Specific Actions

    Toast messages now provide success or failure feedback when users perform actions like submitting feedback or copying content, enhancing the overall user experience.

  • Improved Design for Feedback Modal and Error Banners

    The design of the feedback modal and error banners has been updated to better align with the chat and medical consultation services.

V2.1(Sep 24th, 2024)

  • Improved Korean LLM Model

    The chat service's Korean language capabilities have been further enhanced.

  • Improved Line Spacing in Chat

    Line spacing in chat text has been adjusted to improve readability.

V2.0(Sep 9th, 2024)

  • Launch of Model Hub 2.0

    Model Hub 2.0 has been officially launched.