Skip to main content

Account Engines

This page lets you set your personal model preferences that apply across your entire account. Whatever you set here will override the engine settings inside individual characters and individual servers. How to get here:
In the left sidebar, click Account → then click Engines
The full path is: Account → Engines Think of it like this: you have engine settings in three places — at the character level, the server level, and the account level. Account-level settings win. They have the highest priority and override everything else.
Priority order (highest to lowest):
  1. Account Engines (this page) — overrides everything below
  2. Server Dedicated Engines (per-server settings in Discord integration) — overrides character defaults
  3. Character AI Engine (Universal → AI Engine inside each character) — only applies if neither of the above is set

AI Engine

Full path: Account → Engines → AI Engine section This section is at the top of the Engines page. It controls which AI language model (the “brain”) all your characters use by default.

Primary AI Engine

SettingWhat it does
Primary AI Engine (dropdown)The main AI model used to generate responses for all your characters. Click the dropdown to see available models and select one.
Next to the dropdown, there may be three icons:
  • Wrench icon 🔧 — click this to open extra settings for the selected model
  • Info icon ℹ️ — click this to read about what this model is good at
  • Clear icon ✕ — click this to reset back to the default model

Fallback AI Engine

SettingWhat it does
Fallback AI Engine (dropdown)The backup AI model that activates automatically if the primary engine goes down or returns an error. This keeps your characters online even during model outages.
Choose a fast, reliable model as your fallback — something lightweight like Gemma 3n. It doesn’t need to be the most powerful model, just one that is stable and responds quickly so your characters never go fully offline.

Voice Configuration

Full path: Account → Engines → Voice Configuration section This section has two separate settings — one for the voice itself (how it sounds) and one for the AI that decides what to say during voice calls.

Voice Engine

SettingWhat it does
Voice Engine (dropdown)The text-to-speech (TTS) model that converts text into spoken audio for your character. This controls what your character sounds like. Example: Eleven V3 by ElevenLabs.

Voice Processing Model

SettingWhat it does
Voice Processing Model (dropdown)The AI language model specifically used to write what your character says during voice calls. This is separate from the voice synthesis — it decides the content, not the sound.
Next to each voice processing model option, you’ll see its estimated latency (e.g., 1.1s). Pick a model with lower latency for smoother, faster voice conversations.
Two different settings, two different jobs:
  • Voice Engine = controls HOW the character sounds (the voice/audio quality)
  • Voice Processing Model = controls WHAT the character says during voice calls (the AI writing the words)
Both can be set independently.

Image Generation

Full path: Account → Engines → Image Generation section This section controls which image model is used when any of your characters generates an image.

Image Engine

SettingWhat it does
Image Engine (dropdown)The AI image generation model used account-wide for all image generation. Example: Gemini 3 Pro.
Select from the available models in the dropdown. This setting applies to all your characters unless you’ve configured a different model inside a specific character’s Universal → Image Engine settings — though that character-level setting will be overridden by this account-level setting if both are set.

Configuration Priority — Quick Reference

PriorityWhere the setting livesPath to get thereOverrides
1 (Highest)Account EnginesAccount → EnginesEverything
2Server Dedicated EnginesServer integration settingsCharacter-level defaults
3 (Lowest)Character AI EngineUniversal → AI Engine → AI EnginesNothing (default only)
Use the Account Engines page to set a consistent baseline that applies to all your characters. Only configure engine settings inside individual characters when you specifically need one character to use a different model than the rest.