souvik18/Roy-v1

TEXT GENERATIONConcurrency Cost:1Model Size:7BQuant:FP8Ctx Length:4kPublished:Dec 20, 2025License:apache-2.0Architecture:Transformer Open Weights Cold

Roy-v1 is a 7 billion parameter LLaMA-based personal AI assistant model developed and fine-tuned by Souvik Pramanick. Designed for conversational interaction, it excels at general assistance, problem-solving, coding support, and content generation. This model is optimized for practical everyday tasks, offering helpful and friendly responses.

Loading preview...

Roy-v1: A Personal AI Assistant

Roy-v1 is a 7 billion parameter LLaMA-based AI assistant model, independently developed and fine-tuned by Souvik Pramanick. It is designed to be a helpful, conversational, and practical tool for a wide range of everyday tasks.

Key Capabilities

  • Natural Conversation: Engages in fluid and helpful dialogue.
  • Knowledge & Problem Solving: Answers general knowledge questions and solves mathematical or logical problems.
  • Coding Assistance: Aids in coding tasks, including debugging and generating code snippets.
  • Content Generation: Capable of writing emails, stories, and various forms of content.
  • Learning Support: Explains complex concepts simply and assists with brainstorming.

Good For

Roy-v1 is ideal for users seeking a versatile AI assistant for personal use, learning, and productivity. Its LLaMA-based architecture and 7B parameters make it suitable for deployment on various systems, including those with limited resources, especially with community-provided quantized versions. The model is openly available for community usage.

Finetunes
1 models