v0.1.1

Features

• Support base64 audio embeddings requests (#44).

Python binding

• Return embedding vectors as lists without NumPy (#43).

v0.1.0

First release

clap.cpp provides C++17/ggml audio and text embedding inference with GGUF models.

Highlights

• Model support: CLAP, CLaMP3, MERT v1, MERT v2 FullSong, MuQ, and MuQ-MuLan.
• WAV, MP3, and FLAC decoding with automatic downmixing and resampling.
• clap embedding for native text/audio embeddings and clap serve for an OpenAI-compatible text embeddings API.
• CPU Python, Rust, and Node.js bindings with self-contained source distributions.
• Cross-platform CLI release builds for CPU, CUDA, Vulkan, and ROCm targets.