Qwen3-Coder-Next Technical Report
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908858600914944 |
|---|---|
| author | Cao, Ruisheng Chen, Mouxiang Chen, Jiawei Cui, Zeyu Feng, Yunlong Hui, Binyuan Jing, Yuheng Li, Kaixin Li, Mingze Lin, Junyang Ma, Zeyao Shum, Kashun Wang, Xuwu Wei, Jinxi Yang, Jiaxi Zhang, Jiajun Zhang, Lei Zhang, Zongmeng Zhao, Wenting Zhou, Fan |
| author_facet | Cao, Ruisheng Chen, Mouxiang Chen, Jiawei Cui, Zeyu Feng, Yunlong Hui, Binyuan Jing, Yuheng Li, Kaixin Li, Mingze Lin, Junyang Ma, Zeyao Shum, Kashun Wang, Xuwu Wei, Jinxi Yang, Jiaxi Zhang, Jiajun Zhang, Lei Zhang, Zongmeng Zhao, Wenting Zhou, Fan |
| contents | We present Qwen3-Coder-Next, an open-weight language model specialized for coding agents. Qwen3-Coder-Next is an 80-billion-parameter model that activates only 3 billion parameters during inference, enabling strong coding capability with efficient inference. In this work, we explore how far strong training recipes can push the capability limits of models with small parameter footprints. To achieve this, we perform agentic training through large-scale synthesis of verifiable coding tasks paired with executable environments, allowing learning directly from environment feedback via mid-training and reinforcement learning. Across agent-centric benchmarks including SWE-Bench and Terminal-Bench, Qwen3-Coder-Next achieves competitive performance relative to its active parameter count. We release both base and instruction-tuned open-weight versions to support research and real-world coding agent development. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_00729 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Qwen3-Coder-Next Technical Report Cao, Ruisheng Chen, Mouxiang Chen, Jiawei Cui, Zeyu Feng, Yunlong Hui, Binyuan Jing, Yuheng Li, Kaixin Li, Mingze Lin, Junyang Ma, Zeyao Shum, Kashun Wang, Xuwu Wei, Jinxi Yang, Jiaxi Zhang, Jiajun Zhang, Lei Zhang, Zongmeng Zhao, Wenting Zhou, Fan Computation and Language We present Qwen3-Coder-Next, an open-weight language model specialized for coding agents. Qwen3-Coder-Next is an 80-billion-parameter model that activates only 3 billion parameters during inference, enabling strong coding capability with efficient inference. In this work, we explore how far strong training recipes can push the capability limits of models with small parameter footprints. To achieve this, we perform agentic training through large-scale synthesis of verifiable coding tasks paired with executable environments, allowing learning directly from environment feedback via mid-training and reinforcement learning. Across agent-centric benchmarks including SWE-Bench and Terminal-Bench, Qwen3-Coder-Next achieves competitive performance relative to its active parameter count. We release both base and instruction-tuned open-weight versions to support research and real-world coding agent development. |
| title | Qwen3-Coder-Next Technical Report |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2603.00729 |