asr-services

高性能 Linux & Mac 离线中文语音输入法，基于 Ali FunASR. ~0.1s 瞬时上屏，输入法级稳定性， 极高中文准确率、低资源占用(CPU Only).支持 IBus / Fcitx5 / Tahoe 26

本地优先的视频→结构化笔记服务：B站/抖音/本地视频，平台字幕+本地离线转写，LLM 生成带时间轴笔记；桌面应用 / MCP / Agent Skill 三种接入。

A Wyoming protocol ASR proxy that verifies speaker identity and isolates voice commands from background noise before forwarding audio to a downstream speech-to-text service. Designed for Home Assistant voice pipelines to prevent false activations from TVs, radios, and other people - and to deliver clean transcripts even in noisy environments.

Privacy-first AI interview assistant with live transcription, real-time AI suggestions for any type of interviews and coding challenges

自托管 ASR 推理服务

Construct voice from web-videos and then clone it!

open-source speech AI platform for organizations that cannot send sensitive conversations to a third party

Enterprise-grade distributed microservices ASR system. Features dynamic scaling and high-concurrency resilience. Built on Event-Driven Architecture with Go, Redis Streams & Python.

Built on Mega-ASR, a finetune of the qwen asr model that aims to decipher information in high-noise environemnts that suits better irl situations. This is a high-performance inference framework that is designed to maximize speed and is inspired by the insanely-fast-whisper.

🗣 Verify speaker identity and clean voice audio for accurate speech-to-text in Home Assistant using Wyoming protocol ASR proxy.

A lightweight, CPU-based ASR service using the SenseVoice model. Built with FastAPI and optimized for deployment via Docker to cloud platforms.

Offline speech recognition for robot Agents on NVIDIA Jetson. Fast, private, real-time voice input over DDS. Tested on Unitree G1 and Galbot G1.

🎧 X2AGI speech services: ASR, diarization, AI reports (gRPC, REST clients)

Advanced Cyber Speech Intelligence Platform powered by Faster-Whisper & NLP. Features real-time transcription, toxicity detection, speaker diarization, and batch processing.

Streaming platform with real-time captions

Server to run ASR (particularly Whisper and Qwen-ASR) for project Hermes

Local, offline push-to-talk dictation for macOS — hybrid speech recognition (GigaAM + Whisper), live captions, global hotkeys, no cloud

Routines to convert MMIF transcripts into other formats and create transcript metadata

MeetCap is a self-hosted Discord meeting assistant for a small internal development team.