Qwen Code v0.12.4-preview.0 Released
💡New /review skill + fixes boost Qwen Code for AI coding agents
⚡ 30-Second TL;DR
What Changed
Added bundled /review skill for out-of-the-box code review
Why It Matters
Enhances developer productivity with ready-to-use code review and robust fixes, reducing errors in AI-assisted coding workflows. Welcomes new contributors, signaling growing community support for Qwen Code agents.
What To Do Next
Upgrade to v0.12.4-preview.0 and test the /review skill for automated code reviews in your projects.
Key Points
- •Added bundled /review skill for out-of-the-box code review
- •Fixed LLM failures in qualitative insights and added deepseek-r1 token limits
- •Reject PDF files to prevent session corruption
- •Improved shell tool truncation and rapid pty output capture
- •Disabled symlinks on Windows to fix git clone install failures
🧠 Deep Insight
Background and context from public sources — not the original article. 9 sources cited.
🔑 Enhanced Key Takeaways
- •Qwen Code is an open-source terminal-based AI agent optimized for Qwen3-Coder models, providing agentic workflows with built-in tools and subagents for understanding large codebases[1][6][8].
- •It supports integration with IDEs like VS Code, JetBrains, Zed, and ACP, as well as optional plugins and experimental LSP services[1][4][5].
- •Qwen Code is inspired by tools like Claude Code and Google’s Gemini CLI, offering terminal-first experiences with features like concurrent runners and multimodal input support[3][4][6].
📊 Competitor Analysis▸ Show
| Feature | Qwen Code | Cursor | Qoder | DevoxxGenie |
|---|---|---|---|---|
| Primary Interface | Terminal CLI with IDE plugins | Full IDE | Vertically integrated IDE | JetBrains IDE plugin |
| Model Optimization | Qwen3-Coder (480B MoE) | Flexible models incl. Qwen3 | Qwen3-Coder with NES | Various local/cloud models |
| Key Strengths | Agentic workflow, local run, 92 langs | Polished reliability, multi-file | Deep Qwen integration | IDE-native experience |
| Pricing | Free (Apache 2.0) | Subscription-based | Free/open | Free/plugin-based |
| Benchmarks | Sonnet 4.5-level perf w/ 3B active params | High reliability scores | Strong multi-step edits | Good for JetBrains users |
🛠️ Technical Deep Dive
- •Qwen3-Coder-Next, the optimized model for Qwen Code, features 80B total parameters with 3B activated via MoE (512 experts, 10 activated per token, 1 shared)[2][3].
- •Architecture includes hybrid layout: 48 layers with Gated DeltaNet for long-range dependencies, Gated Attention for reasoning, and MoE for efficiency; hidden dimension 2048[2][3].
- •Native 256K-262K context length supports long-horizon reasoning, complex tool use, and recovery from failures; trained with executable task synthesis and RL[2][3].
- •Runs locally on consumer hardware (e.g., 64GB MacBook, RTX 5090) in non-thinking mode without generating
blocks[2][3].
🔮 Future ImplicationsAI analysis grounded in cited sources
⏳ Timeline
📎 Sources (9)
Factual claims are grounded in the sources below. Forward-looking analysis is AI-generated interpretation.
- secondtalent.com — Qwen AI for Coding Reviews Usage and Performance
- Hugging Face — Qwen3 Coder Next
- dev.to — Qwen3 Coder Next the Complete 2026 Guide to Running Powerful AI Coding Agents Locally 1k95
- qwenlm.github.io — Roadmap
- alibabacloud.com — Qwen Code Coding Plan
- mydeveloperplanet.com — Getting Started with Qwen Code for Coding Tasks
- qwen.ai — Qwencode
- GitHub — Qwen Code
- qwen.ai — Blog
Weekly AI Recap
Read this week's curated digest of top AI events →
👉Related Updates
AI-curated news aggregator. All content rights belong to original publishers.
Original source: Qwen (GitHub Releases: qwen-code) ↗
This is a summary, not the original. Read the source, or get the weekly briefing.
Weekly AI briefing
One email a week. Unsubscribe anytime.