transformers-js - Run machine learning models in JavaScript
Runs pre-trained machine learning models in JavaScript or TypeScript for NLP, computer vision, audio, and multimodal tasks.
Tags
Updated: 2026-10-16machine learningJavaScriptTypeScriptnatural language processingcomputer visionaudio processingmultimodal
Typical Inputs
Typical Outputs
What this skill does
- Run machine learning models
- Classify text
- Generate and translate text
- Summarize text
- Answer questions
- Classify images and audio
- Detect objects
- Segment images
- Recognize speech
- Synthesize speech
- Extract image captions
- Extract embeddings
Inputs
- Task name
- Pre-trained model
- Text
- Images
- Audio files
- Document images
- Candidate labels
- Device options
- Quantization options
Outputs
- Classification results
- Generated text
- Translations
- Summaries
- Question answers
- Detected objects
- Image segments
- Depth estimates
- Transcriptions
- Audio output
- Image captions
- Embeddings
Requirements
- Node.js 18 or later
- Modern browser with ES modules
- Compatible WebGPU environment for WebGPU execution
- Internet access for downloading models
- Transformers.js package