LingXizhiIDE
中 / ENGet Started
RUSH 2.0 deep task engine is now live→

AI-Native IDE
Intelligence in every line of code

RUSH multi-phase orchestration · Semantic code graph · 70% token cost savings
Natively supports 15+ leading models including Kimi K3 / GLM-5.2 / DeepSeek — or bring your own API key
Download for FreeView Docs
Windows · macOS · Linux
rush-task.tsauth.gouser.go✦ AI Panel
▾ lingxizhi
rush-task.ts
auth.go
user.go
▾ internal
jwt.go
sso.go
1// RUSH：重构用户认证模块
2import { lxz } from "@lxz/core"
3 
4const task = await lxz.rush({
5  goal: "接入 JWT 刷新 + 单点登录",
6  agents: ["planner", "coder", "verifier"],
7  budget: { tokens: 50_000, cache: "l2" }
8})
9 
10task.on("phase", (p) => console.log(p.name, p.progress)
LingXizhi AI · Kimi K3
Add SSO logic to the refresh endpoint
Generated sso.go and updated two call sites in auth.go ; token refresh now carries the device fingerprint.
+86 −12 lines changed
✓ Accept allReject
✦ RUSH phase 3/6 · ~42s remaining
70%

Token cost savings (three-layer cache)

3.2×

Average productivity boost

15+

Natively supported models

<50ms

FIM completion first-token latency

NATIVE SUPPORT FOR LEADING MODELS

Kimi K3GLM-5.2DeepSeek-V4Qwen3.6Doubao Seed 2.1MiniMax M3VolcengineKimi K3GLM-5.2DeepSeek-V4Qwen3.6Doubao Seed 2.1MiniMax M3Volcengine
FEATURES

A complete toolchain built for AI-native development

From completion to multi-agent orchestration, from local core to cloud sandbox — every stage is redesigned for efficiency and cost.

RUSH Deep Task Engine

A one-sentence requirement is automatically decomposed into a plan-code-verify multi-phase pipeline, with multi-agent parallel execution and quality-gated convergence.

Explore RUSH →

AI Chat & Inline Editing

Multi-turn sidebar chat, inline rewriting of selections, one-click accept with diff preview — Preserved Thinking keeps long tasks on track.

Learn more →

FIM Smart Completion

Whole-line and whole-block completion driven by the semantic code graph, <50ms first-token latency, precise prediction even across files.

Learn more →

Semantic Code Graph

Repo-wide symbol-level indexing and dependency analysis — the AI sees the structure of your entire project, not fragments.

Learn more →

MCP Ecosystem & Tool Market

Native MCP client/server with one-click install of browser, database, and deployment tools — agent capabilities without limits.

Browse market →

Cloud Sandbox & Security

Dangerous commands run isolated, three-level file recovery, AES-256 encrypted key storage — enterprise permissions and auditing out of the box.

Security whitepaper →
RUSH ENGINE

One sentence,
one complete feature delivered

RUSH turns "requirement → code" into an observable, intervenable multi-phase pipeline. Each phase is verified independently; failures roll back and re-plan automatically.
Requirement understanding & task planningThe Planner decomposes natural language into an ordered subtask DAG and assesses the blast radius
Parallel coding in isolationMultiple Coder agents work in parallel inside separate worktrees without interfering with each other
Quality gates & automated verificationThe Verifier runs tests, static checks, and diff reviews — anything below the bar goes back for rework
Converged delivery & experience retentionMerges the best result, generates a change report, and writes lessons learned into long-term memory
rush — zsh
$ lxz rush "重构用户认证模块"
✦ Plan complete · 6 subtasks · 14 files affected
▸ [1/6] Analyze existing auth flow ······ done  1.2s
▸ [2/6] Design JWT refresh scheme ······ done  2.8s
▸ [3/6] Implement refresh endpoint 72%
○ [4/6] SSO integration ······ queued
○ [5/6] Tests & verification ········ queued
○ [6/6] Convergence & report ········ queued
⣿ RUSH running · 12,400 tokens saved (L2 cache hit 83%)
PRICING

Simple pricing, counted by conversation

No hidden token bills. Plan quotas are based on daily Agent runs — clear and predictable. Save ~17% with yearly billing.

Free

For personal trial and light use
¥0/ Free forever
  • 10 Agent runs per day
  • Full IDE basics
  • FIM smart completion
  • Cloud sandbox & sync storage
  • Priority support
Start for Free
Most Popular

Pro

The daily driver for professional developers
¥29/ mo · ¥290 billed yearly
  • 1,000 Agent runs per day
  • RUSH deep task engine
  • 6 cloud sandboxes
  • 5GB config sync storage
  • Priority support
Upgrade Now

Plus

For heavy users and power players
¥99/ mo · ¥990 billed yearly
  • Unlimited Agent runs
  • RUSH engine + advanced features
  • 10 cloud sandboxes
  • 20GB config sync storage
  • Priority ticket support
Contact to Enable
DOWNLOAD

Every platform, ready out of the box

Current version v2.1.0 · Built-in Rust core auto-update · 5-minute first-time setup
Windows
Windows 10+ · x64 · 82 MB
Download for Windows
macOS
macOS 12+ · Apple Silicon / Intel · 88 MB
Download for macOS
Linux
Ubuntu 20.04+ · .deb / .AppImage · 79 MB
Download for Linux
FAQ

Frequently Asked Questions

Which models does LingXizhi support? Can I bring my own API key?
Natively supports 15+ leading models including Kimi K3, GLM-5.2, DeepSeek-V4, Qwen3.6, Doubao, and MiniMax, via official direct connections and aggregators like Volcengine. You can also bind your own API key in Settings — billed by the provider with zero platform markup.
What happens when I use up my daily Agent runs?
Your daily quota resets the next day. If you regularly need more, you can upgrade at any time: Pro offers 1,000 Agent runs per day and Plus is unlimited — upgrades take effect immediately.
Will my code be uploaded to your servers?
No. LingXizhi is local-first: your code stays on your device by default, and only the necessary context is encrypted and sent to your chosen model when you make an AI request. Sensitive data such as API keys is stored with AES-256 encryption, and cloud sandboxes are fully isolated from your local environment.
Can I migrate from VS Code / Cursor?
Yes. A one-click migration wizard on first launch imports your keybindings, theme, settings, and common plugin configuration from VS Code / Cursor — a 5-minute first-time setup and you're up and running.
Limited time: save ~17% with yearly billing

Ready to make development 3.2× faster?

Download LingXizhi IDE now and start AI-native development in 5 minutes.
Download for FreeCreate Account