# xszyou/Fay

fay是一个帮助数字人（2.5d、3d、移动、pc、网页）或大语言模型（openai兼容、deepseek）连通业务系统的agent框架。

Repository: https://github.com/xszyou/Fay
Canonical: https://ross.abutalabs.com/products/fay
Language: Python
License: GPL-3.0
License Family: copyleft
Topics: ai, python, android, api, ue5, unity3d, wav2lip
Last push: 2026-08-07T03:57:47+00:00

## Health v2 (maintenance only)
Score: 75/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 96, release rhythm 35, longevity 100
- inputs: {"age_days": 1542, "days_push": 26, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 13455, forks 2358 (observed 2026-08-28T04:11:03.342797+00:00)

## What it is
Fay is an open-source Python digital human framework that connects 2.5D/3D/mobile/web digital humans or OpenAI-compatible LLMs to business systems. It provides streaming voice and text interaction interfaces with pluggable ASR, TTS, and LLM models, plus agent tool-calling and MCP support.

## Use cases
- build a virtual assistant with voice interaction
- create a digital human live streamer or virtual teacher
- connect an LLM chatbot to my business system
- run an offline voice assistant on a local server
- add speech recognition and TTS to a digital avatar
- build a multi-user concurrent voice agent with custom knowledge base
- integrate a digital human into a website or mobile app
- configure voice commands that trigger actions

## When to choose
- you need a full-stack digital human pipeline with ASR, TTS, LLM, and avatar driving
- you want offline-capable voice assistants with wake word and interruption support
- you need to serve multiple concurrent users from one assistant backend
- you want agent tool calling and MCP integration in a digital human
- you need auto-broadcast mode for virtual anchors or news presenters

## When to avoid
- you only need a simple text chatbot without voice or avatars
- you want a lightweight library to embed rather than a full framework service
- you need a permissively licensed project for closed-source commercial use (GPL-3.0)
- you require production-grade enterprise support and SLAs

## Facets
- artifact type: framework
- maturity: active
- function: agent-framework, speech-recognition, tts, llm-inference, chatbot, rag, mcp, api-framework, http-server
- domain: artificial-intelligence, large-language-models, chatbots, speech-processing, developer-tools
- platform: python, windows, cross-platform, self-hosted
- tags: digital-human, virtual-assistant, voice-assistant, asr, tts, livestreaming, virtual-broadcaster, wake-word, bionic-memory, tool-calling, ai-agents, natural-language-processing, macos, linux, android, web-server

## Member repositories
- xszyou/Fay (main) score 75

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:11:03.342797+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:13:17.526618+00:00, confidence not recorded.
  - readme: https://github.com/xszyou/Fay (fetched 2026-08-28T04:11:03.342797+00:00, sha 25a32c2c4c56)
  - registry_pypi: https://pypi.org/pypi/fay/json (fetched 2026-08-29T08:08:11.286505+00:00, sha a6bdea4236cd)
- Data as of 2026-08-30T08:39:29.467469+00:00.