Skip to content
Agent Radar

Qwen Code 0.27.0 adds parallel child questions and GitHub repo filters

Updated Qwen Code

Summary

Qwen Code v0.27.0 (November 7, 2026) lets managed-agent sessions hold several child questions at once and adds repository allowlists to the GitHub channel editor. Desktop v0.27.0 followed on November 8.

Tool
Qwen Code
Category
Coding agents
Date
Open source
Yes
License
Apache-2.0
Pricing
Not stated in the sources we cited.
Sources

Qwen Code published v0.27.0 on GitHub on November 7, 2026, and Desktop v0.27.0 the next morning (November 8). The notes list no known breaking changes. The CLI and Desktop notes overlap, so they are covered together rather than as separate launches.

Release facts

FactDetail
ProductQwen Code
Versionv0.27.0 (CLI); desktop-v0.27.0
Release dateNovember 7, 2026 (CLI); November 8, 2026 (Desktop)
LicenseApache-2.0, as reported by the GitHub API
Breaking changesNone listed
Where to get itGitHub release v0.27.0

What changed

  • Parallel child questions. Managed-agent sessions can hold several child questions at once, answer them in any order, and keep drafts and errors separated per child (#16484).
  • Web Shell task review. The composer adds a task Q&A heatmap; clicking a day scrolls to that day's first Q&A. Pending Ask User Question panels get an Ask for details action that opens a side task to explain options without answering them (#16037, #16458).
  • GitHub repository filters. The GitHub channel editor adds allowlist and blocklist repository filters, and the notes say bound pull-request delivery and plain comment delivery become more reliable (#16493, #16416).
  • Safer shell edits. Shell tool edits handle CRLF and Unicode in sed substitutions, and cancelled commands report as interrupted while keeping partial output (#15701, #15836).
  • File-change hooks. Local non-managed sessions gain FileChanged hooks for add, change and unlink events on watched files and directories (#13969).
  • Credential handling. Upgrade refusals and Advisor startup errors no longer expose embedded credentials, and empty Harness user info is rejected during start and verify (#16352, #16083).

The notes also say the managed-agent runtime is faster and less prone to deadlocks, with fixes around session rename, snapshot publication, storage admission and lock handling. That is the project's own description, not an independent benchmark.

Who should upgrade now

Teams that run managed-agent sessions with several pending questions, or that attach Qwen Code to many GitHub repositories, get the most from this release: child questions no longer have to finish one at a time, and the new repository filter can keep notifications on an allowlist or blocklist. If you rely on the GitHub channel for PR comments, read the delivery fixes before rolling the update out.

Alternatives

Qwen Code compared with its alternatives
ToolLicensePricing note
Qwen CodeThis launchAn open-source AI coding agent that lives in your terminal to turn ideas into code.
Apache-2.0Not stated
AiderApache-2.0 terminal pair programmer. The project describes Aider as AI pair programming in your terminal.
Apache-2.0Apache-2.0 software. Model access is separate.
ClineApache-2.0 coding agent published as an SDK, IDE extension, or CLI assistant in the cline/cline repository.
Apache-2.0The repository license is Apache-2.0. Model access is separate.
GooseApache-2.0 agent from Block. The repository describes Goose as an open-source, extensible AI agent that goes beyond code suggestions.
Apache-2.0Apache-2.0 client. Provider access is separate.
OpenCodeOpen-source coding agent from the anomalyco/opencode repository, MIT-licensed, with a terminal client and a site at opencode.ai.
MITThe client is MIT-licensed. Model access is separate; this page does not quote a hosted plan.

More from Qwen Code

View all

Questions

What is Qwen Code?

According to the official documentation and repository description, Qwen Code is an open-source agentic coding tool that runs in your terminal and helps turn ideas into code.

Is Qwen Code open source?

Yes, according to the GitHub repository, Qwen Code is open source under the Apache-2.0 license.

Which model providers can Qwen Code use?

According to the overview documentation, Qwen Code can connect to Alibaba ModelStudio (via Token Plan or API key), third-party providers such as DeepSeek, MiniMax, Z.AI, Kimi, and OpenRouter with an API key, or custom local servers and proxies.

What is the latest launch covered here?

Qwen Code 0.27.0, published November 7, 2026, with Desktop v0.27.0 the next day. Managed-agent sessions can hold several child questions at once, and the GitHub channel editor adds repository allowlists and blocklists.

Filed on Agent Radar. Claims in this note are limited to the sources above.