Artificial Intelligence at Göttingen University

AI Tools

Selected AI Tools


The GWDG AI services offer a versatile platform for practical AI applications. The portfolio ranges from a chatbot with advanced capabilities (e.g. retrieval-augmented generation, tool integration, MCP) to speech transcription and image editing.

All services can be conveniently used via a web interface or seamlessly integrated into existing systems via an OpenAI-compatible API.
Developed as part of the KISSKI project (AI Service Center for Sensitive and Critical Infrastructures), these services meet high data protection requirements and are therefore particularly suitable for sensitive application scenarios.

SAIA

Scalable AI accelerator
API access for AI services

SAIA is the scalable AI accelerator that hosts our AI services. These services include Chat AI and CoCo AI. SAIA API keys (application programming interface) can be requested and used to access the services directly from your code.


In brief:

  • Secure, scalable infrastructure
  • Local digital sovereignty
  • Seamless AI integration
Documentation
Link to the tool

Chat AI

Chatbot with multiple LLMs/language models and no data storage

Chat AI offers simple and secure access to powerful generative AI. Via the intuitive interface, users can chat directly with a selection of different AI models. It can be used without data storage.



In brief:

  • Secure, flexible chat
  • Advanced tool integration
  • Personalised chat personas
Documentation
Link to the tool

Coco AI

Programming support through AI
AI code completion in the editor

CoCo AI is a service for code completion with chat AI. To use it, you need a SAIA API key. Many code editors offer LLM integration. We provide documentation on using the Continue extension for Visual Studio Code (VSCode) as well as for the Zed editor.

In brief:

  • AI in everyday development work
  • Full model control
  • Comprehensive AI support
Documentation

Voice AI

Speech-to-text transcription, including live

Voice AI offers an intuitive and secure platform for transcribing and translating audio files. Users can easily upload their audio or video files to have them transcribed in the original language or translated into English.

In brief:

  • Natural transcription
  • Real-time support
  • Reliable speaker recognition
Documentation
Link to the tool

Image AI

Image generation through text input

ImageAI is an AI-based image generation service. The service uses a high-performance computing (HPC) architecture and the FLUX (text-to-image) and Qwen (image-to-image) models to generate images intuitively based on prompts.

In brief:

  • Intelligent image generation
  • Precise image editing

Documentation
Link to the tool

Protein AI

3D protein structure prediction from amino acid sequences (FASTA)

Protein AI enables state-of-the-art protein structure prediction. Using high-performance computing (HPC) and AlphaFold2/Boltz models, our platform delivers fast, precise, and reliable automatic structure predictions for various proteins.

In brief:

  • Protein structure prediction
  • Near-experimental accuracy

Documentation
Link to the tool

Arcana / RAG

Your own documents as an AI knowledge base for targeted answers

Arcana is a retrieval-augmented generation (RAG) service that enables interaction with documents—such as research papers, manuals, or study materials—in natural language. The Arcana service works together with the Chat AI service by using the relevant information from a document collection (Arcana) to generate better answers.

In brief:

  • Access to scientific knowledge
  • Scalable knowledge transfer
  • Academic collaboration platform
Documentation
Link to the tool