AIHubMixAIHubMix
Models
kat-dev
Compare
Qwen logo

Kat Dev

kat-dev
Qwen
KAT-Dev (32B) is an open-source 32B parameter model specifically designed for software engineering tasks. It achieved a 62.4% resolution rate on the SWE-Bench Verified benchmark, ranking fifth among all open-source models of various scales. The model is optimized through multiple stages, including intermediate training, supervised fine-tuning (SFT) and reinforcement fine-tuning (RFT), as well as large-scale agent reinforcement learning (RL). Based on Qwen3-32B, its training process lays the foundation for subsequent fine-tuning and reinforcement learning stages by enhancing fundamental abilities such as tool usage, multi-turn interaction, and instruction following. During the fine-tuning phase, the model not only learns eight carefully curated task types and programming scenarios but also innovatively introduces a reinforcement fine-tuning (RFT) stage guided by human engineer-annotated “teacher trajectories.” The final agent reinforcement learning phase addresses scalability challenges through multi-level prefix caching, entropy-based trajectory pruning, and efficient architecture.

Pricing

  • Input Tokens: $0.137 /M tokens
  • Output Tokens: $0.548 /M tokens

Input Modalities

  • Text

Output Modalities

  • Text

Capabilities

  • Tools

Try this model

Python
import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["AIHUBMIX_API_KEY"],
    base_url="https://aihubmix.com/v1",
)

response = client.chat.completions.create(
    model="kat-dev",
    messages=[
      {
        "role": "user",
        "content": "Hello, how are you?"
      }
    ],
    max_tokens=1024,
    stream=False,
)

print(response.choices[0].message.content)

Frequently asked questions

What is kat-dev?

KAT-Dev (32B) is an open-source 32B parameter model specifically designed for software engineering tasks. It achieved a 62.4% resolution rate on the SWE-Bench Verified benchmark, ranking fifth among all open-source models of various scales. The model is optimized through multiple stages, including intermediate training, supervised fine-tuning (SFT) and reinforcement fine-tuning (RFT), as well as large-scale agent reinforcement learning (RL). Based on Qwen3-32B, its training process lays the foundation for subsequent fine-tuning and reinforcement learning stages by enhancing fundamental abilities such as tool usage, multi-turn interaction, and instruction following. During the fine-tuning phase, the model not only learns eight carefully curated task types and programming scenarios but also innovatively introduces a reinforcement fine-tuning (RFT) stage guided by human engineer-annotated “teacher trajectories.” The final agent reinforcement learning phase addresses scalability challenges through multi-level prefix caching, entropy-based trajectory pruning, and efficient architecture.
AIHubMix

© 2023 - 2026 AIHubMix, LLC

Product

Company

Developer

Connect