hamzah0asadullah/TinyRP-0.8B
TEXT GENERATIONConcurrency Cost:1Model Size:0.8BQuant:BF16Ctx Length:32kPublished:Feb 17, 2026License:apache-2.0Architecture:Transformer Open Weights Cold

TinyRP-0.8B by hamzah0asadullah is a 0.8 billion parameter language model based on Qwen3-0.6B, designed to provide usable role-playing abilities while retaining general-assistant qualities. This model is optimized for efficient execution on mobile devices and laptops, making it suitable for on-device applications. It enhances role-playing capabilities while preserving the reasoning abilities of its base model, offering a balance between specialized function and broad utility.

Loading preview...

TinyRP-0.8B: Role-Playing and General Assistance for Small Devices

The hamzah0asadullah/TinyRP-0.8B model is part of the TinyRP series, an initiative to bring effective role-playing capabilities to smaller language models without sacrificing their general-assistant functions. Built upon the Qwen3-0.6B base, this 0.8 billion parameter model is specifically designed for efficient operation on resource-constrained devices like mobile phones and laptops.

Key Capabilities

  • Enhanced Role-Playing: TinyRP-0.8B significantly improves role-playing interactions, allowing for more immersive and character-consistent dialogues, as demonstrated by detailed examples in the model's documentation.
  • Preserved Reasoning: Despite its small size and role-playing focus, the model maintains the reasoning capabilities inherited from its Qwen3-0.6B base, enabling it to handle tasks such as mathematical problem-solving.
  • Efficient Performance: Optimized for smaller devices, it offers acceptable speeds for on-device inference, making it a practical choice for local applications.

Recommended Usage

For optimal performance, specific inference parameters are suggested, including a temperature of 0.6 (higher for non-scientific tasks), Top-P of 0.95, and Top-K between 9 and 21. The model's ability to perform both complex role-playing and accurate mathematical calculations makes it a versatile option for developers looking for a compact yet capable LLM.