Casual-Autopsy/The-True-Abomination-24B

TEXT GENERATIONPricing:Input $0.7 / Cached $0.04 / Output $1.16Concurrent Unit Cost:2Model Size:24BQuant:FP8Context Size:32kTool Calling:SupportedPublished:May 12, 2025Architecture:Transformer0.0K Featherless Exclusive Cold

The-True-Abomination-24B by Casual-Autopsy is a 24 billion parameter merged language model with a 32768 token context length, specifically designed for enhanced compatibility with SillyTavern features and extensions. It excels in roleplay scenarios and improves reasoning capabilities, particularly with legacy reasoning methods like Status block/thinking box CoT. This model is optimized for immersive roleplaying experiences where traditional reasoning prompts are preferred over modern, more explicit CoT techniques.

Loading preview...

The-True-Abomination-24B: Roleplay Optimized Merge

This 24 billion parameter model, developed by Casual-Autopsy, is a sophisticated merge designed to maximize compatibility with SillyTavern features and extensions while maintaining strong roleplaying capabilities. Built upon a base of TheDrummer/Cydonia-24B-v2 and incorporating numerous other models, it leverages advanced merge methods like SCE, Della, and CABS.

Key Capabilities

  • SillyTavern Compatibility: Engineered for seamless integration with various SillyTavern features and extensions, enhancing the user experience for interactive storytelling.
  • Enhanced Roleplay: Specifically tuned to perform well in diverse roleplaying scenarios, providing coherent and engaging narrative contributions.
  • Legacy Reasoning Support: Improves the model's ability to handle 'legacy reasoning' (e.g., Status block/thinking box CoT), which is preferred for less immersion-breaking thought processes. Users can boost this by setting max reasoning prompts to 3+ or injecting CoT formatting prompts.

Good For

  • SillyTavern Users: Ideal for those seeking a robust model that works exceptionally well within the SillyTavern ecosystem.
  • Immersive Roleplaying: Excellent for generating detailed and consistent roleplay responses, especially when a more subtle approach to reasoning is desired.
  • Developers Experimenting with Merges: Showcases the application of SCE, Della, and CABS merge methods for specialized model creation.