StudyArenanew chat

Aion-RP 1.0 (8B)

Small & localOpen weightsText8B parameters

An open Llama 3.1 8B fine-tune for character-driven roleplay.

This model competes in the arena. See where it ranks.

About the Aion-RP family

Aion-RP is Aion Labs' downloadable Llama 3.1 8B fine-tune, trained for character fidelity and roleplaying dialogue. Unlike the newer Aion systems, its OpenRouter endpoint exposes plain text generation without reasoning controls or tools.

All Aion-RP models

How this one is configured

This configuration has ~8B parameters and a 32K-token context window.

Who makes it

Aion LabsPolandCreative-model lab specializing in collaborative roleplaying and storytelling systems.

Student board

← Back to all models