FundamentalsUpdated 2026-10-23
Multimodal
A model that can process or generate multiple modalities such as text, images, audio and video.
#多模态#图像#视频#语音
In plain language
Multimodal models can answer questions about images, write code from screenshots, or generate images and video from text. Support varies by model; check input types, resolution and output formats. Multimodal capabilities often combine with OCR, speech recognition and video generation.
Example
Upload a financial report screenshot and ask the model to extract the table and explain key metrics.
Recommended tools
Related terms
FundamentalsFeatured
Large Language Model
A model trained on massive text to understand and generate natural language.
3 related terms
Applications
Optical Character Recognition
Converting text in images or scans into editable text.
3 related terms
ApplicationsFeatured
Text-to-Video
Generating video clips from text descriptions.
3 related terms
Applications
Agentic Workflow
Breaking a task into steps planned by AI, executed by tools and reviewed by humans.
3 related terms