open-yoki

open-yoki is an open-source AI live-stream host that runs voice, a real-time digital face, chat replies, memory and posting on a single server. The GitHub repo jacke-pp/open-yoki was created on October 28, 2026 (~340★). It is the base stack behind the commercial yoki host; the open build omits the original persona, voice and face.

Repo: jacke-pp/open-yoki · Commercial context: superidollive.com · Maker: @Jacke_pp

Key points

1. Product shape: one-box livestream AI host (voice + face + chat + memory + social posts), aimed at RTMP destinations such as Binance Square.
2. Open vs commercial: MIT open-yoki is the runnable base; branded yoki assets stay closed.
3. Freshness: repo created Oct 28, 2026 with a same-week star spike, matching this scan window's GitHub fresh cut.

Caveats

Coverage is mostly GitHub/CN chatter rather than a large English press launch—score reflects a distinctive OSS name with traction, not a mega-launch.

How to try it

1. Clone jacke-pp/open-yoki and follow the README for single-server setup.
2. Point an RTMP output at your streaming destination; bring your own voice/face assets if you need a branded host.
3. Contact the maker via X (@Jacke_pp) for commercial yoki questions.