Qwen Code
Alibaba's free, open-source terminal coding agent, Gemini CLI fork
Overview
Qwen Code is Alibaba's free, open-source (Apache 2.0) terminal coding agent — a fork of Google's Gemini CLI rebuilt around Qwen3-Coder — that plugs into OpenAI, Anthropic, Gemini or Alibaba Cloud backends. Built for developers who want a scriptable CLI agent, not a GUI IDE.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 12 Jul 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
What you can produce with Qwen Code
- Free, Apache 2.0 open-source CLI (26k GitHub stars)
- Terminal UI plus VS Code, Zed and JetBrains integrations
- Multi-provider auth: OpenAI, Anthropic, Gemini, Alibaba DashScope, OpenRouter, Fireworks, local models
- Auto-memory, auto-skills, sub-agents and agent-team modes
- MCP (Model Context Protocol) support out of the box
- Headless/scripting mode for CI and automation
- Built-in code-review and debugging skills
ASEAN Perspective
Qwen Code in Southeast Asia
ASEAN-region availability and pricing notes coming soon. Drop the editorial team a note via /contact/ if you can supply local context (Singapore/Malaysia/Indonesia/Thailand/Vietnam).
Qwen Code is what you'd expect from an open-source fork of Gemini CLI retooled for Qwen3-Coder: the same terminal-first workflow and project-context system, now provider-agnostic across OpenAI, Anthropic, Gemini and Alibaba's own DashScope API. Qwen3-Coder scores 69.6% on SWE-bench Verified, ahead of the Gemini 2.5 Pro it was forked from and within range of Claude Sonnet-tier models, which makes the free CLI genuinely competitive rather than a budget compromise.
The rough edge is cost predictability: Alibaba killed the free OAuth tier in April 2026, so getting good results now means either bringing your own API key (OpenRouter still lists a free qwen3-coder tier) or paying $50/month for Alibaba Cloud's Pro Coding Plan. Best for developers comfortable in a terminal who want model flexibility without vendor lock-in; GUI-first coders should look at Trae or Cursor instead.
What people say
The repository — github.com/QwenLM/qwen-code — carries 26,000 GitHub stars and 2,600 forks, with TypeScript making up over 80% of the codebase, evidence of an active open-source project rather than a marketing shell for Alibaba's models.
Independent benchmarking puts Qwen3-Coder, the model Qwen Code is built around, at 69.6% on SWE-bench Verified — ahead of Gemini 2.5 Pro's 63.8% (the model the tool was forked from) and close to Claude-tier results, though writeups comparing it directly to Claude Code note Claude still wins on autonomous multi-file coordination and has a more mature MCP tool ecosystem. For developers who want a free, competitive terminal agent, reviewers consistently point to Qwen Code over paid alternatives; for professional agentic workflows where reliability matters more than cost, the same reviewers still recommend paying for Claude Code.
The free ride got shorter in 2026. Qwen's OAuth free tier — which let anyone run the CLI against Alibaba's hosted Qwen3-Coder at no cost — was discontinued on April 15, 2026. Coverage of the shutdown points users toward OpenRouter's qwen/qwen3-coder:free endpoint or Fireworks AI as no-cost fallbacks, both usable as bring-your-own-key backends since Qwen Code is provider-agnostic by design. Alibaba's own paid path, the Cloud Coding Plan, now has a single Pro tier at $50/month (90,000 requests/month, capped at 6,000 per 5-hour window) since the cheaper Lite tier stopped accepting new subscribers on March 20, 2026.
Feature-wise, the tool ships auto-memory, auto-skills, sub-agents and agent-team modes, MCP support out of the box, and integrations for VS Code, Zed and JetBrains alongside the core terminal UI — a broader surface than a typical CLI wrapper. Because it's Apache 2.0 and model-agnostic, nothing stops you from running it against a completely different model provider, which is the strongest argument for it over closed competitors: no lock-in, and the code itself is auditable.
Summary of public user & expert reviews, compiled by RECATOOLS.
About this listing
This entry was compiled from publicly available data including Qwen Code's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with Qwen Code unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to Qwen Code directly →
Spotted something out of date? Suggest an update →
Qwen Code in the news
More in Code & Dev Tools
