IT Center Changes

Bug fixes and improvements:

• Improved OpenAI compatibility of the /chat/completions endpoint
• Revised and improved examples in the API documentation
• Optimized overall API performance
• Fixed an issue where the reasoning output was not displayed in some cases
• Increased the timeout for image generation to reduce interruptions during longer generation processes

Features: 

• With this release, RWTHgpt now supports reasoning.
• If multiple Reasoning Effort values are available, users can select them directly in the frontend.
• Where available, reasoning content is displayed in the frontend.

Bug Fixes and Improvements:

• Improved processing of tool calls in the Responses endpoint.

Bugfixes and Improvements:

• Improved the overall modal design, especially for mobile devices
• Enhanced the design of the error modal
• Free models can now continue to be used even after the hard limit has been reached
• Fixed an issue where chat titles would sometimes only appear after reloading the page

Features:

• For employees and students, the model GPT-5.5 has been added. Data processing for GPT-5.5 takes place exclusively within the European Union.

• For employees, the following usage limits apply:
  • GPT-5.5 (Reasoning high): 100 requests per 3 hours
  • GPT-5.5 (Reasoning minimal): 100 requests per hour

• For students, the following usage limits apply:
  • GPT-5.5 (Reasoning high): 100 requests per day
  • GPT-5.5 (Reasoning minimal): 100 requests per 6 hours

• For GPT-5.4-mini, the usage limits have been adjusted as follows:
  • Employees: 1,000 requests per hour
  • Students: 250 requests per 6 hours

• The following models have been removed:
  • GPT-5.2
  • GPT-5.3

Bugfixes and Improvements:

• The response endpoint now returns an error message when the context window is exceeded

Bugfixes and Improvements:

• An issue with permission checks for creating shared chats has been fixed.

• API access to selected RWTHgpt models is now available for employees without additional usage limitations:
  • gpt-oss-120b
  • e5-mistral-7b-instruct
  • qwen3-embedding-8b
  • mistralai-mistral-small-4-119b

Bugfixes and Improvements:

• Opening a chat no longer triggers a full page reload
• For OpenAI models hosted via GWDG, only the supported chat completions endpoint is now available for selection
• Text overflow issues with long words and URLs have been fixed
• The allowed values for reasoning_effort have been corrected for certain OpenAI models
• Chat links can now be opened directly even when the user is not logged in
• Shared chats can now be opened directly even when the user is not logged in

Features:

• API access to selected RWTHgpt models is now available for students without additional usage limitations:
  • gpt-oss-120b
  • e5-mistral-7b-instruct
  • qwen3-embedding-8b
  • mistralai-mistral-small-4-119b