Kimi Dev 72B

🚀 Kimi-Dev

Introducing Kimi-Dev: A Strong and Open-source Coding LLM for Issue Resolution

Kimi-Dev-72B is a newly developed open-source coding large language model (LLM) designed for software engineering tasks. It has achieved a new state-of-the-art performance on SWE-bench Verified among open-source models, demonstrating its strong capabilities in issue resolution.

Introducing Kimi-Dev:
A Strong and Open-source Coding LLM for Issue Resolution

Kimi-Dev Team

🚀 Quick Start

from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "moonshotai/Kimi-Dev-72B"

model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype="auto",
    device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained(model_name)

prompt = "Give me a short introduction to large language model."
messages = [
    {"role": "system", "content": "You are a helpful assistant."},
    {"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)

generated_ids = model.generate(
    **model_inputs,
    max_new_tokens=512
)
generated_ids = [
    output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
]

response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]

✨ Features

• Outstanding Performance: Kimi-Dev-72B achieves 60.4% performance on SWE-bench Verified, surpassing the runner-up and setting a new state-of-the-art result among open-source models.
• Optimized via Reinforcement Learning: It is optimized through large-scale reinforcement learning. It can autonomously patch real repositories in Docker and only gain rewards when the entire test suite passes, ensuring correct and robust solutions that meet real-world development standards.
• Easy to Access: Kimi-Dev-72B is available for download and deployment on Hugging Face and GitHub, welcoming developers and researchers to explore its capabilities and contribute to its development.

Performance of Open-source Models on SWE-bench Verified.

📄 License

This project is licensed under the MIT license.

📚 Documentation

Model Information

📝 Citation

@misc{kimi_dev_72b_2025,
  title        = {Introducing Kimi-Dev: A Strong and Open-source Coding LLM for Issue Resolution},
  author       = {{Kimi-Dev Team}},
  year         = {2025},
  month        = {June},
  url          = {\url{https://www.moonshot.cn/Kimi-Dev}}
}