Model reference · open weights

Kimi-Dev

LLMs moonshotai Text gen 1 build Open weights 1k dl/mo

Kimi-Dev is an open-weight language model from moonshotai. Kimi-Dev-72B (BF16) weighs 145 GB; the smallest configuration that runs it is 4× L40S 48 GB.

  • Kimi-Dev-72B is an open-source coding large language model developed by moonshotai for software engineering tasks, specifically issue resolution.
  • The model features 72.7 billion parameters and supports a context length of 131,072 tokens.
  • It is released under the MIT license and achieves 60.4% performance on the SWE-bench Verified benchmark.

Summary of the moonshotai/Kimi-Dev-72B model card, 2026-10-01

What it is

Released bymoonshotai
Released2025-06-16
Parameters72.7B
VRAM145 GB for the weights

What it runs on

Memory and cards for Kimi-Dev-72B (BF16)

145 GBweights, file size
328 MBcache per 1K tokens
962 MBruntime overhead, at least
131,072 tokenscontext max
CardRequests at onceContext maxMemory
8K each32K each
RTX 3060 12 GB … H200 141 GB
11 smaller cards
———
B200 180 GB8269K176 GB
4× L40S 48 GB
tensor parallel
9279K44.0 GB a card
2× RTX PRO 6000 Blackwell 96 GB
tensor parallel
11290K93.8 GB a card
2× H200 141 GB
tensor parallel
4411all 128K138 GB a card
4× H100 80 GB
tensor parallel
5313all 128K78.1 GB a card
4× A100 80 GB
tensor parallel
6015all 128K78.2 GB a card
2× B200 180 GB
tensor parallel
7117all 128K176 GB a card
Memory needed at each load
Requests at once8K tokens each32K tokens each
1149 GB157 GB
5160 GB200 GB
8168 GB232 GB
16189 GB318 GB
32232 GB490 GB
64318 GB834 GB

One card, with vLLM's small-card settings.

From the model card

What moonshotai says about Kimi-Dev

Read the model card

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

We introduce Kimi-Dev-72B, our new open-source coding LLM for software engineering tasks. Kimi-Dev-72B achieves a new state-of-the-art on SWE-bench Verified among open-source models.

  • Kimi-Dev-72B achieves 60.4% performance on SWE-bench Verified. It surpasses the runner-up, setting a new state-of-the-art result among open-source models.

  • Kimi-Dev-72B is optimized via large-scale reinforcement learning. It autonomously patches real repositories in Docker and gains rewards only when the entire test suite passes. This ensures correct and robust solutions, aligning with real-world development standards.

  • Kimi-Dev-72B is available for download and deployment on Hugging Face and GitHub. We welcome developers and researchers to explore its capabilities and contribute to development.

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]

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}}
}

Quoted from the model card on Hugging Face. The full card is behind the Hugging Face link above.

© 2026 AxForge · EU-hosted AI infrastructure Pricing Docs Trust Privacy Terms