it0is0me/kanana-1.5-8b-instruct-2505-Safe-DPO

TEXT GENERATIONPricing:Input $0.37 / Cached $0.074 / Output $0.38Concurrent Unit Cost:1Model Size:8BQuant:FP8Context Size:8kTool Calling:SupportedPublished:Aug 14, 2026Architecture:Transformer Featherless Exclusive Cold

The it0is0me/kanana-1.5-8b-instruct-2505-Safe-DPO is an 8 billion parameter instruction-tuned language model. This model is designed for general conversational AI tasks, focusing on safe and aligned responses through Direct Preference Optimization (DPO). It is suitable for applications requiring a balanced and moderated interaction, leveraging its 8192 token context length for coherent and extended dialogues.

Loading preview...

Model Overview

The it0is0me/kanana-1.5-8b-instruct-2505-Safe-DPO is an 8 billion parameter instruction-tuned language model. This model has been developed with a focus on safety and alignment, utilizing Direct Preference Optimization (DPO) during its training process to ensure moderated and appropriate responses. It is designed to handle a wide range of conversational tasks, providing a robust foundation for various AI applications.

Key Capabilities

  • Instruction Following: Capable of understanding and executing instructions provided in natural language.
  • Safe and Aligned Responses: Optimized through DPO to generate outputs that are considered safe and aligned with desired behavioral norms.
  • General Conversational AI: Suitable for broad conversational use cases, from chatbots to interactive assistants.
  • Extended Context Handling: Features an 8192 token context length, allowing for more coherent and contextually aware interactions over longer dialogues.

Good For

  • Applications requiring a general-purpose conversational AI.
  • Use cases where safety and alignment of generated text are critical.
  • Developing chatbots or virtual assistants that need to maintain consistent and moderated interactions.
  • Scenarios benefiting from a model with a reasonable context window for sustained conversations.