---
title: SWE-Dev-7B
canonical_url: "https://www.modelscope.cn/models/ZhipuAI/SWE-Dev-7B"
md_url: "https://www.modelscope.cn/models/ZhipuAI/SWE-Dev-7B.md"
repository: ZhipuAI/SWE-Dev-7B
last_updated: 2026-01-20
license: mit
model_type:
  - qwen2
architectures:
  - Qwen2ForCausalLM
parameters: 545.0M
tensor_type:
  - BF16
library_name:
  - safetensors
  - pytorch
frameworks:
  - pytorch
inference_backends:
  - "deploy_task text/emb"
  - "lmdeploy 0.9.1"
  - "lmdeploy_turbomind 0.9.1"
  - "sglang 0.5.2"
  - "vllm 0.9.2"
downloads: 85
stars: 0
---

# SWE-Dev-7B

> SWE-Dev-7B - ZhipuAI 在 ModelScope 开源的模型。📝 Paper | 🌐 Github

ZhipuAI/SWE-Dev-7B 是 ModelScope 魔搭社区上的 545.0M 参数机器学习模型，采用 mit 许可，可用 deploy_task text/emb、lmdeploy 0.9.1、lmdeploy_turbomind 0.9.1 部署。

- **Repository**: ZhipuAI/SWE-Dev-7B
- **License**: mit
- **Parameters**: 545.0M
- **Inference backends**: deploy_task text/emb, lmdeploy 0.9.1, lmdeploy_turbomind 0.9.1, sglang 0.5.2, vllm 0.9.2
- **Downloads**: 85
- **Stars**: 0
- **Last updated**: 2026-01-20

Source: https://www.modelscope.cn/models/ZhipuAI/SWE-Dev-7B

---

📝 [Paper](https://arxiv.org/abs/2506.07636) | 🌐 [Github](https://github.com/THUDM/SWE-Dev/)

- 🤗 [SWE-Dev-7B (Qwen-2.5-Coder-7B-Instruct)](https://huggingface.co/THUDM/SWE-Dev-7B/)
- 🤗 [SWE-Dev-9B (GLM-4-9B-Chat)](https://huggingface.co/THUDM/SWE-Dev-9B/)
- 🤗 [SWE-Dev-32B (Qwen-2.5-Coder-32B-Instruct)](https://huggingface.co/THUDM/SWE-Dev-32B/)
- 🤗 [SWE-Dev-train (Training Data)](https://huggingface.co/datasets/THUDM/SWE-Dev-train/)

🚀 SWE-Dev, an open-source Agent for Software Engineering tasks! This repository contains the SWE-Dev-32B model as presented in the paper [SWE-Dev: Building Software Engineering Agents with Training and Inference Scaling](https://huggingface.co/papers/2506.07636).

💡 We develop a comprehensive pipeline for creating developer-oriented datasets from GitHub repositories, including issue tracking, code localization, test case generation, and evaluation.

🔧 Based on open-source frameworks (OpenHands) and models, SWE-Dev-7B and 32B achieved solve rates of 23.4% and 36.6% on SWE-bench-Verified, respectively, even approaching the performance of GPT-4o. 

📚 We find that training data scaling and inference scaling can both effectively boost the performance of models on SWE-bench. Moreover, higher data quality further improves this trend when combined with reinforcement fine-tuning (RFT). For inference scaling specifically, the solve rate on SWE-Dev increased from 34.0% at 30 rounds to 36.6% at 75 rounds.