missang/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 missang/kanana-1.5-8b-instruct-2505-Safe-DPO model is an 8 billion parameter instruction-tuned language model developed by missang. This model is designed for general-purpose conversational AI, focusing on safety and adherence to instructions through Direct Preference Optimization (DPO). With an 8192-token context length, it is suitable for a variety of interactive text generation tasks where safe and aligned responses are critical.

Loading preview...

Model Overview

The missang/kanana-1.5-8b-instruct-2505-Safe-DPO is an 8 billion parameter instruction-tuned language model developed by missang. It is designed to provide safe and aligned responses, leveraging Direct Preference Optimization (DPO) during its training process. The model supports a context length of 8192 tokens, making it capable of handling moderately long inputs and generating coherent, extended outputs.

Key Characteristics

  • Parameter Count: 8 billion parameters, offering a balance between performance and computational efficiency.
  • Instruction-Tuned: Optimized to follow user instructions effectively, enhancing its utility in conversational and task-oriented applications.
  • Safety-Focused: Incorporates Direct Preference Optimization (DPO) to improve safety and reduce undesirable outputs.
  • Context Length: Features an 8192-token context window, allowing for more detailed conversations and processing of longer documents.

Potential Use Cases

  • General Conversational AI: Suitable for chatbots and virtual assistants requiring instruction adherence and safe responses.
  • Content Generation: Can be used for generating various forms of text content, from creative writing to informative summaries.
  • Instruction Following: Excels in tasks where precise adherence to user prompts is crucial.
  • Safe AI Applications: Ideal for deployments where mitigating harmful or biased outputs is a primary concern.