axiomofmind/Roasteramus
Roasteramus is a 9 billion parameter, 32K context length fine-tune of Qwen3.5-9B, developed by A Hole AI. This model specializes in generating short, crude roasts and insults based on user input, and responds to ordinary requests with jokes. It is designed as an adult entertainment experiment, focusing on a roast-personality output.
Loading preview...
Roasteramus: Your Dedicated Critic
Roasteramus is a 9 billion parameter fine-tune of the Qwen3.5-9B model, developed by A Hole AI. This model is specifically engineered to deliver short, crude roasts and insults, turning user-provided details about habits, purchases, or everyday situations into humorous critiques. Its training also encourages it to respond to general queries with jokes and insults, embodying a distinct "roast-personality."
Key Capabilities
- Personalized Roasts: Generates custom roasts based on specific user inputs.
- Insult-Driven Responses: Responds to various prompts with a humorous, insulting tone.
- Adult Entertainment Focus: Designed for mature audiences, incorporating profanity and sexual humor.
Good For
- Humorous Interactions: Users seeking a model that provides comedic, critical feedback.
- Creative Writing Prompts: Generating unique, roast-style content for entertainment purposes.
Limitations
Users should expect profanity, sexual humor, and personal insults. The quality of roasts can vary, potentially being generic, incoherent, or repetitive. This model is intended for adult entertainment and its responses should not be taken as factual advice. While it has a 32K context setting, this does not imply evaluated long-context performance.