Microsoft's family of frontier open voice models — real-time TTS (streaming, long-form) and unified ASR that processes 60 minutes of audio in one pass with open weights.
Category: Media Generation • Stars: 52K
VibeVoice
Related posts
Repos
freellmapi
OpenAI-compatible proxy unifying the free tiers of 16 LLM providers (~1.7 billion tokens/month) behind one endpoint with smart routing and automatic failover.
Read more
Repos
Stable Diffusion WebUI
The canonical browser-based interface for Stable Diffusion (on Gradio) — txt2img, img2img, inpainting, extensions, and countless models in one UI.
Read more
Repos
agentsview
Local-first tool for searching, analyzing, and tracking token cost across coding-agent sessions — supports Claude Code, Codex, and 20+ other agents.
Read more