Models
Notis uses a variety of specialized AI models to deliver different functionalities across the platform. Each model is optimized for specific tasks to ensure the best performance and user experience.
Triage Engine Model - gpt-5
Acts as the intelligent router that determines which model or response type is most appropriate for each user request. This ensures you always get the most relevant and accurate assistance.
Purpose: Triaging user requests and determining appropriate responses
Assistant Model - gpt-5
The primary model powering general assistant interactions and queries throughout Notis. This model handles most user conversations and provides comprehensive responses to a wide range of questions.
Purpose: Main model for assistant interactions and general queries
Web Search Model - gpt-4o-search-preview
Specialized for web search capabilities and browsing functionality. This model helps find and process information from across the internet to supplement responses.
Purpose: Web search and browsing capabilities
Deep Research Model - o4-mini-deep-research
Specialized model designed for conducting thorough research and complex analysis tasks. When you need in-depth information or comprehensive analysis, this model takes over to provide detailed insights.
Purpose: In-depth research and analysis tasks
Image Model - gpt-image-1
Handles all image-related processing, including image analysis, description, and visual content understanding. This model can interpret charts, diagrams, photos, and other visual content.
Purpose: Image processing and analysis
Merge Model - gpt-5
Responsible for consolidating and merging multiple messages or information sources into coherent responses. This ensures smooth conversation flow and organized information delivery.
Purpose: Merging and consolidating messages
Speech Model - whisper-1
Powers speech-to-text functionality, converting audio input into text that can be processed by other models. Essential for voice interactions and audio content transcription.
Purpose: Speech-to-text transcription
Last Updated: August 4, 2025
Text-to-Speech Model - gpt-4o-mini-tts
Converts text responses into natural-sounding speech output. This model enables audio playback of responses and voice-based interactions.
Purpose: Text-to-speech generation
TTS Preprocessing Model - gpt-4.1
Prepares and optimizes text content before it's converted to speech by the TTS model. This ensures high-quality audio output with proper pronunciation and flow.
Purpose: Preprocessing text before text-to-speech conversion
Model Updates
Notis regularly updates its AI models to improve performance, accuracy, and capabilities. The models are maintained and updated based on:
Model updates are automatically applied to ensure you always have access to the latest AI capabilities. No action is required on your part when models are updated.