CrispASR
A single C++ binary offering 119 speech backends, including 62 text-to-speech engines, for automatic speech recognition, text-to-speech, forced alignment, and translation, with no Python dependency at runtime.
CrispASR is a unified speech engine, started as a fork of whisper.cpp, that bundles full ggml C++ runtimes for a wide range of open-weight speech recognition and text-to-speech architectures into one build and one consistent command-line interface. It picks the right backend for a given model file automatically, or lets the backend be chosen explicitly, and needs no Python, PyTorch, or per-model binary to run, only the compiled binary and a GGUF model file.
Every backend compiles to WebAssembly as well, so it can run entirely client-side in a browser, and a hosted demo shows live transcription, text-to-speech, and language detection. Companion projects in the same ecosystem include a Flutter transcription app built on CrispASR, a related engine for embeddings, retrieval, and OCR, and a Python GUI covering a similar set of speech backends.
It exposes a command-line interface, an HTTP server, a C ABI, and bindings for Python, Rust, Dart, Go, Ruby, and Java.