audio-ai-hub

Why we included this project

Anyone tracking speech and audio machine learning will find this hub a practical shortcut: it collects papers, open model weights, benchmarks, and datasets in one place, filterable by category, star count, and date. Coverage spans audio LLMs, speech recognition, text-to-speech, and music generation, and the featured tier alone shows real range, from Whisper for recognition to CosyVoice and IndexTTS for synthesis and MusicGen for generation. Star counts refresh weekly and submissions are curated, so the list stays current instead of turning into a pile of stale bookmarks. If your job is choosing which open audio model to try next, this is a faster way to get oriented than working through raw search results.