mucgpt
Features
This page highlights the key capabilities of MUCGPT. Future work is listed in the Roadmap.
MUCGPT is an assistant-first product. Assistants are reusable, workflow-specific chat configurations with their own instructions, behavior, optional tools, and sharing scope. Chat is where work happens; assistants provide structure so recurring work becomes faster and more reliable.
Chat
The chat provides a clear space to work with the underlying language model on almost any topic.
- Configure a system instruction to guide responses and adjust temperature to balance factual vs. creative output.
- Multi-turn conversations are supported. History stays local in the browser (IndexedDB). You can continue past conversations and mark favorites for quick access.
Create Assistants
Design assistants that encode recurring instructions, tone, constraints, and model settings. Optionally add starter prompts, follow-up actions, and tools. Private assistants remain yours; you decide when to share more broadly.
Share Assistants
Share assistants with selected teams or wider audiences. MUCGPT makes ownership, scope, and configuration transparent so colleagues understand what an assistant does before relying on it.
Extensible Tools
Assistants can use built-in and extendable tools. Examples include:
- Summarization for condensing long texts.
- Brainstorming to generate mind maps that can be exported and refined.
- Easy Language to translate complex text into simpler language.
These examples are illustrative—the platform is designed to grow with additional tools as organizational needs evolve.
AI Governance And Compliance Checks
Assistant system prompts can be screened against configured EU AI Act high-risk use cases before they are saved. High-risk findings route the immutable assistant version to an administrator/legal review queue. See the AI governance and compliance checks guide for flow diagrams, configuration, verification rules, and the review lifecycle.
Start Page
A focused entry point to discover assistants, jump back into recent work, and access tutorials.
Model Selection
Choose from available models and adjust creativity to fit the task. Model choices are transparent to the user.
Speech-to-Text (Beta)
Dictate directly into the chat input: a microphone button appears next to the input field once speech-to-text is enabled. Transcribed text is appended to what you have already typed, so typing and dictation can be freely combined.
- Runs entirely on your device. Audio is transcribed locally in your browser — no microphone input leaves your device. The resulting transcript is placed in the chat input and can be submitted through the normal chat request.
- Model choice with language hints. The settings dialog lists every model with its download size and supported languages, so you can pick the right one at a glance. Models download once and are cached in the browser.
- Language follows the interface language. German interface → German transcription, with automatic language detection where the model supports it.
- Voice activity detection. Recording stops automatically after a pause; segments are transcribed live while you speak.
Available models (lineup grows over time):
| Model | Languages | Requirements |
|---|---|---|
| Whisper Small | Multilingual (incl. German) | None (runs on CPU/WASM) |
| Whisper Large v3 Turbo | Multilingual (incl. German) | WebGPU |
| Whisper Large v3 Turbo German | German | WebGPU |
| Distil-Whisper Large v3 German | German | None (runs on CPU/WASM) |
| NVIDIA Canary 180M Flash | English, German, French, Spanish | None (runs on CPU/WASM) |
| NVIDIA Parakeet TDT 0.6B v3 | 25 languages, automatic detection (incl. Ukrainian) | None (runs on CPU/WASM) |
The feature must be enabled per deployment; if the microphone button is missing, your administrator has not activated it yet.
Tutorials
Guided examples and tips to help first-time and experienced users get value quickly.
Dark Mode and i18n
Accessible UI with dark mode support and internationalization.
General Features
Generated text in various formats is displayed correctly. The formats currently supported are:
- Markdown
- PLAIN HTML
- Mermaid diagrams embedded in Markdown code blocks
- Mathematical formulas written in LaTeX, enclosed within
$$ ... $$delimiters