Qoder
Alibaba's agentic IDE: chat-based Agent mode, autonomous Quest tasks
Overview
Qoder is Alibaba's agentic coding IDE, built around a Repo Wiki that maps a whole codebase and a Quest mode that runs multi-file tasks unattended. It routes across Claude, GPT, Gemini and Qwen, and moved from free preview to credit-based paid tiers in 2026.
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 These figures reflect RECATOOLS research as of 12 Jul 2026 and may be out of date or incomplete.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.
- Basic model access
- Limited chat/agent messages
- 2,000 credits/month
- Code completion + chat/agent
- Quest, Repo Wiki, Knowledge Card
- 6,000 credits/month
- Higher resource allocation
- 20,000 credits/month
- Maximum resource allocation
- 2,000 credits/seat/month
- Prorated seat billing
What you can produce with Qoder
- Repo Wiki auto-generated codebase map
- Quest mode: autonomous multi-file task delivery
- Agent mode chat-based pair programming
- Routes across Claude, GPT, Gemini and Qwen models
- Qoder CLI for terminal-based agent workflows
- GitHub Action for CI-triggered coding tasks
- Supports 200+ programming languages
ASEAN Perspective
Qoder 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).
The closest thing China has to an internationally marketed agentic IDE, Qoder answers Cursor and Windsurf with an English-first site, multi-model routing across Claude, GPT, Gemini and Qwen, and a genuinely different Quest mode that takes a written spec and delivers a tested multi-file change without hand-holding. The Repo Wiki, which builds a navigable map of a codebase's architecture and dependencies, is the feature reviewers keep citing as the standout.
The rough edges are real: indexing a medium-sized repo can take hours before Repo Wiki pages are usable, and Chat mode tends to rewrite more of a file than a targeted fix needs — one reviewer noted GPT-5 in VS Code left a smaller diff on the same bug. Auto-indexing caps out around 10,000 files, ruling out very large monorepos. Best fit: solo developers and small teams on greenfield or mid-size codebases who want to delegate whole tasks, not just get better autocomplete.
What people say
Launched into public preview in August 2025, Qoder is positioned as the international, standalone evolution of Alibaba's earlier Lingma coding assistant — a full IDE (an editor, not a plugin) built around what Alibaba calls context engineering: a hybrid retrieval system combining vector search, code graphs and a pre-built Repo Wiki that maps a project's architecture before you start prompting.
Developer Jimmy Song's hands-on review is the most detailed independent account so far. He found the Repo Wiki genuinely useful for onboarding to unfamiliar code, but painfully slow to build on a medium-sized project — "only 5% of the index after 2 hours," with some pages remaining external links rather than locally cached content even after three hours. He also noted Qoder's Chat mode tends to over-edit: on the same bug fix, GPT-5 in VS Code produced a tighter, more targeted diff. Hard limits exist too — Repo Wiki tops out around 6,000 files and auto-indexing caps at 10,000, putting large monorepos out of reach.
Quest mode is the more distinctive feature: hand it a written spec and it works asynchronously, planning, coding and testing on its own, only pinging you when it's blocked or needs a decision. Coverage on Hacker News and dev blogs describes it as a hybrid of Kiro's spec-first workflow and Cursor's editing feel, with one reviewer calling well-written specs the single biggest lever on Quest's output quality.
Qoder moved from free preview to a credit-based subscription in 2026: Pro at $20/month (2,000 credits), Pro+ at $60 (6,000 credits) aimed at heavy Agent users, and Ultra at $200 (20,000 credits) for teams running long Quest tasks regularly, plus a free Community tier and a Teams plan starting at $30/seat for two or more seats. Alibaba has also cut list prices on its own Qwen3.7 models by up to 80% for off-peak international users, undercutting Claude- and GPT-based competitors on raw token cost even when Qoder is routing to those same third-party models.
Trade coverage frames it as competitive for repo-aware, multi-file refactors and end-to-end feature delivery, but weaker than Cursor for fast, tab-completion-heavy workflows — a tool built for delegating a task, not for staying in constant keystroke-level flow.
Summary of public user & expert reviews, compiled by RECATOOLS.
About this listing
This entry was compiled from publicly available data including Qoder's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with Qoder 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 Qoder directly →
Spotted something out of date? Suggest an update →
More in Code & Dev Tools
