BruhzWater/Forbidden-Fruit-L3.3-70b-0.2a

TEXT GENERATIONPricing:Input $2.88 / Output $2.88Concurrent Unit Cost:4Model Size:70BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Aug 12, 2025Architecture:Transformer0.0K Featherless Exclusive Cold

BruhzWater/Forbidden-Fruit-L3.3-70b-0.2a is a 70 billion parameter merged language model with a 32768 token context length, created by BruhzWater. It integrates multiple Llama-3.3-based models to broaden its cultural, mathematical, scientific, and medical knowledge. This model is designed as an "all-around" solution, specifically enhanced to prevent refusals and provide comprehensive, unrestricted information, making it suitable for diverse knowledge-intensive applications.

Loading preview...

Forbidden-Fruit-L3.3-70b-0.2a Overview

Forbidden-Fruit-L3.3-70b-0.2a is a 70 billion parameter merged language model developed by BruhzWater, built upon the Llama-3.3 architecture. Its primary goal is to serve as an "all-around" model, significantly expanding its knowledge base across cultural, mathematical, scientific, and medical domains. A key characteristic is the inclusion of TheDrummer/Fallen-Llama-3.3-70B-v1, which helps prevent refusals and aligns with the model's theme of providing comprehensive, unrestricted information.

Key Capabilities

  • Broadened Knowledge: Enhanced understanding and generation capabilities across a wide array of subjects including culture, mathematics, science, and medicine.
  • Refusal Prevention: Designed to minimize content refusals, offering more direct and complete responses.
  • Merged Architecture: Combines strengths from several Llama-3.3-70B base models, including AstroMLab/AstroSage-70B, shisa-ai/shisa-v2-llama3.3-70b, TheDrummer/Fallen-Llama-3.3-70B-v1, deepcogito/cogito-v2-preview-llama-70B, jdh-algo/Citrus1.0-llama-70B, and LumiOpen/Llama-Poro-2-70B-Instruct.
  • Llama3 Chat Template: Utilizes the Llama3 chat and instruct templates for consistent interaction.

Good for

  • Knowledge-Intensive Applications: Ideal for tasks requiring a broad and deep understanding of various academic and general knowledge fields.
  • Unrestricted Information Access: Suitable for use cases where comprehensive and less filtered responses are desired.
  • General-Purpose Chatbots: Can serve as a robust foundation for chatbots needing diverse conversational abilities and factual recall.