Livetoon Launches Tatara AI Character Creation Service
Tatara moves AI characters from a catalog users pick from to a tool users build themselves, opening the character creation pipeline Livetoon kept internal in kaiwa to outside creators.
Key Facts
- Livetoon Inc. launched Tatara on September 1, 2026, lifting the invitation-only system that had run since July 29, 2026.
- Tatara lets users load VRM models they hold usage rights to and set voice, personality, tone, name, and role to create desktop AI characters.
- Tatara integrates with OpenAI Codex, GitHub Copilot, and OpenCode for coding tasks.
- The Standard plan costs 1,500 yen per month excluding tax, or 1,650 yen including tax, and covers TTS, STT, and LLM conversation functions.
- Tatara runs on Windows 11 (x64 and Arm64) and macOS 13 or later on Apple Silicon; Intel Macs are not supported.
Reporting from 2 sources: ASCII.jp, PANORA.
Livetoon Inc. officially launched Tatara on September 1, 2026, a desktop service that lets users create AI characters from VRM models they hold usage rights to. Users set voice, personality, tone, name, and role, and the character converses by voice, changes facial expressions and motions, and integrates with OpenAI Codex, GitHub Copilot, and OpenCode to handle coding tasks. The invitation-only system that ran since July 29 was lifted, so anyone can download the desktop app from the official website. Tatara builds on technology from Livetoon's earlier voice dialogue service kaiwa, including speech recognition, LLM, speech synthesis, and 3D model expression and motion synchronization. The Standard plan costs 1,500 yen per month excluding tax, or 1,650 yen including tax, and covers conversation functions (TTS, STT, LLM). Free use covers character creation, the desktop app, and forum browsing. Supported environments are Windows 11 (x64 and Arm64) and macOS 13 or later on Apple Silicon; Intel Macs are not supported. Early users during the invitation period built a Windows timer app with their characters, searched for anime, and organized data.
Livetoon, founded in March 2024 and based in Chuo-ku, Tokyo, developed Tatara on the technology behind its earlier voice dialogue service kaiwa, which delivered pre-made 3D AI characters to users. Tatara opens that pipeline so creators set the voice, personality, tone, name, and role themselves. The company's stated direction is to let characters leave the desktop: future plans include exporting characters to streaming, video, and websites, plus a marketplace where voices, personality settings, motions, and plugins can be listed and purchased, with value returning to creators.
During the invitation period, early users reported building a small Windows timer app while consulting their character, checking the finished app and requesting fixes by voice, searching for this season's anime with the character, planning streaming schedules, and having the character add tasks to an external to-do list app. COO Taiga Koshiba said participation during the invitation period exceeded expectations, and development manager Atsuya Tsuchikawa said forum feedback drove feature additions that shaped the release. CEO Kyosuke Kinoshita said Tatara grew from the question of whether characters that were created or purchased could speak, move, and support their owners in daily life, rather than staying limited to streaming and VR spaces.
Synthesized by Yomimono from the 2 cited sources below, including Japanese-language reporting where cited, then editorially reviewed before publishing.