MODELS
laion/clap-htsat-fused
by LAION
Overview
LAION CLAP model checkpoint for multimodal audio-and-language representation and audio-classification workflows.
Details
laion/clap-htsat-fused is a LAION model checkpoint listed on Hugging Face for audio-classification. Hugging Face Transformers documents CLAP as a multimodal audio-and-language model and includes examples that instantiate ClapAudioModel and ClapProcessor from laion/clap-htsat-fused. The official LAION-AI CLAP repository describes the project as providing audio and text representations via Contrastive Language-Audio Pretraining and links the implementation to LAION and the CLAP paper.
When to Use
Use when you need a CLAP checkpoint from LAION for audio-and-language representation tasks. Use with Hugging Face Transformers examples that load ClapAudioModel and ClapProcessor from laion/clap-htsat-fused. Evaluate for audio-classification pipelines where a Hugging Face-hosted CLAP model checkpoint is appropriate.
Getting Started
- Open the Hugging Face model page at https://huggingface.co/laion/clap-htsat-fused to review the model card and files.
- Read the Hugging Face Transformers CLAP documentation for usage patterns with ClapAudioModel and ClapProcessor.
- Review the LAION-AI/CLAP GitHub repository for project context
- implementation details
- and links to the CLAP paper.
- Run a small audio-and-language or audio-classification evaluation before relying on the checkpoint in a production workflow.
Key Features
- •Hugging Face model checkpoint published under laion/clap-htsat-fused.
- •Associated with CLAP
- •described by Hugging Face Transformers as a multimodal audio-and-language model.
- •Supported in Transformers documentation through examples using ClapAudioModel and ClapProcessor.
- •Connected to LAION-AI's CLAP repository for Contrastive Language-Audio Pretraining.
Capabilities
- •audio-classification
- •audio-and-language modeling
- •audio representation
- •text representation
- •contrastive language-audio pretraining
Last updated Jun 4, 2026