Qwen Code CLI: Pricing, Features, Pros & Cons (2026)

Free and open-source agent, with model usage billed through an Alibaba Cloud Coding Plan, Token Plan, or your own API key

What is Qwen Code?

Qwen Code is an open-source AI coding agent for the terminal from Alibaba's Qwen team, licensed under Apache 2.0. It started as a fork of Google's Gemini CLI and has since grown into its own project, built around Qwen models while also connecting to OpenAI, Anthropic, and Gemini APIs, other third-party providers, and local models through Ollama or vLLM.

It extends through MCP servers and built-in skills like /review and /bugfix, and subagents and agent teams split larger tasks across parallel workers. A headless mode runs it inside scripts and CI, a sandbox and Git worktrees isolate changes, and plugins bring it into VS Code, Zed, and JetBrains IDEs.

The agent itself is free, but Qwen discontinued its OAuth free tier on April 15, 2026, so model usage now runs through Alibaba Cloud's Coding Plan subscription, its usage-based Token Plan, or an API key from another provider. Qwen Code doesn't include its own observability or evals, so tracing what each session did and what it cost comes from other tools.

Key features

Core capabilities this platform advertises.

Terminal coding agent with file, shell, and Git tools
Qwen models plus OpenAI, Anthropic, Gemini, and local models
MCP servers and built-in skills
Subagents and agent teams

Strengths and tradeoffs

What this tool does well, and the limitations to keep in mind.

Pros

Apache 2.0 license, free to use with your own API key
Works with Qwen, OpenAI, Anthropic, Gemini, and local models
Headless mode for scripts and CI

Cons

The Qwen OAuth free tier was discontinued in April 2026
The Coding Plan's Lite tier no longer accepts new orders

Plans & pricing

What's included in each plan, and how the tiers compare.

Open source (bring your own key)

Free

The full agent under Apache 2.0, with model usage billed by your chosen provider

Token Plan

TBD

Usage-based

Alibaba Cloud Model Studio models billed per token instead of a fixed monthly fee

Coding Plan Pro

$50

Per month

Common use cases

Developers who want an open-source terminal coding agent that runs Qwen models or models from other providers

Coding in the terminal with Qwen models
Running one agent across several model providers
Automating code review and bug fixes in CI

Using Qwen Code with Respan

Qwen Code routes through the Respan gateway with three OpenAI-compatible environment variables and a Respan API key. Every call is logged and priced automatically, and changing one variable switches Qwen Code to any of 1,000+ models, including Claude, GPT, and Gemini.