Aniimo AI: What the NVIDIA ACE Companion Feature Does
Aniimo uses generative AI through an NVIDIA ACE companion system due in early 2027 on RTX GPUs, limited to your Homeland and answering with in-game actions.
Aniimo’s AI is an NVIDIA ACE companion system, planned for early 2027 and running on RTX GPUs, letting you talk to your own Aniimo in natural language or by voice. Three limits matter: it only works inside your own Homeland, the Aniimo only perform a short list of allowed actions, and responses use in-game actions and emotes rather than generated voice lines.
What the system does
| Feature | Detail |
|---|---|
| Natural language and voice | Talk or type to your own Aniimo |
| Teaching skills | Activities such as crafting and mining can be taught |
| Gathering errands | Send an Aniimo out to collect resources |
| Dynamic emotions | Emotional responses that follow what you do |
NVIDIA’s senior product marketing manager Ike Nnoli put it this way at a pre-Gamescom briefing: “In early 2027, Nvidia ACE will be running on RTX GPUs and it will allow players to speak naturally to their Aniimos. You can chat with them, you can teach them new skills like crafting and mining, and you can discover their distinct personalities.”
Why it is confined to the Homeland
The developers deliberately put the AI in a small box. In the same Q&A, Nnoli explained that developers keep full control over how characters respond, including action and response guardrails and model instructions, and that “specifically for the Aniimo integration, Pawprint have decided to limit it to your homeland, which is your home base where you will do crafting, mining and resource gathering. [It’s also] where the Aniimos themselves can only perform select activities.”
The consequences:
- The AI cannot touch progression, combat or exploration, because those scenes do not open dialogue at all.
- It cannot be talked into skipping quests or handing out rewards, because the executable action set is a short predefined list.
- The Aniimo that keep you company at home are not the same ones you take into the field.
Responses are not generated speech
The system does not produce voiced lines. Aniimo answer through specific in-game actions and emotes instead. The game’s visual language already carries expression, and real-time voice synthesis would not fit it.
What is known and what is not
- Announced: the ACE partnership, early 2027, the RTX requirement, the Homeland restriction, action-and-emote responses, teachable skills and gathering.
- Not announced: a minimum RTX model, language support, whether an extra download is needed, and whether it affects achievements.
Treat it as an enhancement to the home experience, not a system that changes your power.
The Homeland and RV loop is on the gameplay page and multiplayer is on the multiplayer page.