Omo R1

01 / What it was

Omo R1 is an agent who logs into your single-player world as a real player. You talk to him out loud, he answers in his own voice, and then he goes and does it.

Three parts make him: a realtime voice model that listens and speaks, a Hermes agent that plans, and a Mineflayer body that serves 22 tools to that agent. He had his first live run on the 27 October stream and was released as open source the next day.

02 / How it worked

• Voice: a local web page holds a full-duplex voice call. When you ask him to do something, the request goes to his own Hermes profile.
• Body: a Mineflayer bot joins over the local network and serves its actions as tools. The skills are Mindcraft's library, wrapped in jobs that can be interrupted, paused for a fight and resumed.
• Short requests such as "come here", "follow me", "stop", "fight" and "chop some wood" skip the language model.
• Release v0.1.0 ships with 18 tests.

03 / What we saw

Asked for a stone sword, he went from wood to planks, sticks, a crafting table and a pickaxe, then stone and the sword, "including retrying a craft that failed the first time". He was also fragile: he would not stay put, kept dying to zombies, and a request for 8 birch logs ended with 1.

04 / Where it led

The questions he raised (how often a long chain finishes, and how to hear you in game) are in the robots paper.

Sources

1. Omo, brought to life: README. github.com/omoresearch/omo-brought-to-life On stream he ran that whole chain himself, including retrying a craft that failed the first time.
2. Omo, brought to life: CHANGELOG, v0.1.0. github.com/omoresearch/omo-brought-to-life/blob/main/CHANGELOG.md The first public release of Omo as one repo: voice, brain, body and the mod changes that let him in.
3. Stream analysis. stream transcript analysis, 27 Oct 2026 Stay sit sit puppy sit right not working