Current scope
It is not a system that can never give an inappropriate reply. The conversation AI has usage fees, and running it uses electricity. A recording of a full live run has not been published.
Closed source
Open Noa’s YouTube channelNoa.
Comments, answered in voice and expression.
The reply, the voice and the look are separate systems. Noa joins them and answers stream comments with voice and expression.
What it does
- Comment
- judgement
- voice and expression
(03) Noa — FilmIntroduction film (about 35 s · silent · recreated screens and staging)
(+) In detail
07 / CHARACTER & LIVE / Streaming
Meet Noa, an AI gamer and streaming character
The game’s on. Say hello.
Noa plays games and responds to stream comments with her voice and expressions. Drop into a stream, leave a comment, and be part of the conversation.
The source is not public. This link opens Noa’s YouTube channel.
FIRST THING TO TRYOpen Noa’s YouTube channel to explore streams and videos.
Setup, scope & license
Closed source · It is not a system that can never give an inappropriate reply. The conversation AI has usage fees, and running it uses electricity. A recording of a full live run has not been published.
WHAT THE FILM SHOWS
From your comment to Noa’s response.
Reply
Jev judges the comment and decides the reply
Voice
The reply is read out in the Kohaku voice
Look
Image layers are switched to move the face
TRY IT
Explore Noa’s streams and videos.
The channel is available below. Direct links to the explainer (5:48) and introduction film (0:30) will be added once confirmed. Both films include staging.
CURRENT BOUNDARIES
Availability and development.
(05) What it can do
A streaming character who answers comments with a voice and a face.
Source: the private repository · checked on 2026-10-02
Playing the game, answering the chat
Game control runs separately from comment replies, so Noa can respond while playing. The game situation helps determine which line she speaks next. Duplicate and stale comments are filtered out, and questions with fixed answers use prepared lines.
The voice is made on the local PC
Replies are read out with the AivisSpeech voice Kohaku. Speech synthesis has no per-call fee, but the conversation AI has usage fees and the PC uses electricity.
Mouth and eyes follow the voice
Four mouth shapes, blinking and gaze change with the volume of the voice as it plays. Six effects are chosen from the reply, with no extra AI call.
(06) Scope and evidence
What it cannot do,
stated up front.
Public sources checked 2026-09-15. Check each repository and product page for current terms.
SCOPE
It is not a system that can never give an inappropriate reply. The conversation AI has usage fees, and running it uses electricity. A recording of a full live run has not been published.
PROOF / PUBLISHED EVIDENCE
A film made to explain Noa; not a recording of a live stream.
(07) Build with it