cyber-xiaoya-live

cyber-xiaoya-live is an agent app that runs a talking digital human on live streams, answering viewer comments and performing actions for streamers.

What it is good for

  • Running a digital human host on a live stream
  • Replying to viewer comments with a synthesized voice automatically
  • Taking song requests during a livestream
  • Making an avatar perform actions when viewers send commands

At a glance

Who it is for
Setup difficulty
Technical ยท Guided command-line setup

Level 4 of 5. You copy and paste commands into a terminal by following the instructions. Comfortable territory for vibe coders.

Cost
Free and open source
Type
Agent app
Tasks
Works with
Vidu, DeepSeek, OmniVoice, Douyin, Bilibili, YouTube

How to get started with cyber-xiaoya-live

  1. Register a Vidu account, add credit, create an avatar, and generate an API key.
  2. Get an LLM API key, such as DeepSeek, and save it in a text file.
  3. Install Python 3.12 and the dependencies, then copy config.example.json to config.json.
  4. Set up OmniVoice text-to-speech in WSL2, or skip it for a silent rehearsal.
  5. Run .venv\Scripts\python.exe -m src.main, open the stage page, and click the connect button.

From the project's own documentation. If a step does not match what you see, the official pages have the latest instructions.

Before you try it

  • Needs Windows 10/11 and an NVIDIA graphics card with at least 8GB of video memory.
  • Vidu charges 1 credit per second while connected, and a balance under 45 credits blocks starting.
  • You must supply your own avatar, voice, persona, and API keys; none are included.
  • Setup takes about an hour, and text-to-speech is the step most likely to fail.

Momentum

Rank this week
Not ranked (last #8, week 45, 2026)
Best rank
#8 week 45, 2026
GitHub stars
886

GitHub numbers as of Nov 16, 2026. Updated weekly.

429 675 920 Nov 2, 2026 Nov 16, 2026
GitHub stars for cyber-xiaoya-live: 463 on Nov 2, 2026, 886 on Nov 16, 2026. Source: GitHub.

Ranking history

First ranked in week 45, 2026.

Questions

Is cyber-xiaoya-live free?
Yes. It is free and open source (AGPL-3.0 license).
Do I need to know how to code to use cyber-xiaoya-live?
You do not need to write code, but setup means copying and pasting commands into a terminal by following its instructions.
What does cyber-xiaoya-live work with?
According to its own documentation, it works with Vidu, DeepSeek, OmniVoice, Douyin, Bilibili, and YouTube.
How do I get started with cyber-xiaoya-live?
1. Register a Vidu account, add credit, create an avatar, and generate an API key. 2. Get an LLM API key, such as DeepSeek, and save it in a text file. 3. Install Python 3.12 and the dependencies, then copy config.example.json to config.json. 4. Set up OmniVoice text-to-speech in WSL2, or skip it for a silent rehearsal. 5. Run .venv\Scripts\python.exe -m src.main, open the stage page, and click the connect button.
What should I know before trying cyber-xiaoya-live?
Needs Windows 10/11 and an NVIDIA graphics card with at least 8GB of video memory. Vidu charges 1 credit per second while connected, and a balance under 45 credits blocks starting. You must supply your own avatar, voice, persona, and API keys; none are included. Setup takes about an hour, and text-to-speech is the step most likely to fail.
What is cyber-xiaoya-live good for?
Running a digital human host on a live stream, replying to viewer comments with a synthesized voice automatically, taking song requests during a livestream, and making an avatar perform actions when viewers send commands.

Similar agents to cyber-xiaoya-live

  • LocalMiniDrama

    Makes short drama videos from written stories on your computer instead of running a live digital human host.

  • MoneyPrinterTurbo

    Produces finished short videos from a topic, not a live interactive avatar.

  • xiaozhi-esp32

    Voice assistant firmware for ESP32 hardware, not a streaming avatar.

  • bili-cli

    Command-line tool for an agent to act on Bilibili, with no avatar or live host.

  • Voicebox

    Voicebox is a local voice studio that clones voices, generates speech, and dictates text for creators and anyone who wants offline voice tools.

  • VideoCaptioner

    VideoCaptioner is a desktop and command-line tool that adds subtitles to videos, then cleans up and translates them, for people who publish video.

Last updated Nov 16, 2026. Tracked since Nov 2, 2026. Parts of this summary required judgment because the source material was brief. Spotted an error? Tell us.