Code Editors & IDEs, according to AI?
The public record of what ChatGPT, Claude, Gemini, and Perplexity recommend. Ranked across 10 brands, updated weekly.
Model APIAll 4 models put Cursor at #1.
| Consensus rankAPI + Search: measured on the official model API with web search enabled | ChatGPT#1 Cursor | Claude#1 Cursor | Gemini#1 Cursor | Perplexity#1 Cursor |
|---|---|---|---|---|
1 CursorNew entry this week 25 | #10New entry this week | #10New entry this week | #1100New entry this week | #10New entry this week |
2 Visual Studio CodeNew entry this week 21 | #10New entry this week | #10New entry this week | #285New entry this week | #10New entry this week |
3 JetBrains IDEs (IntelliJ, PyCharm, WebStorm)New entry this week 18 | #10New entry this week | #10New entry this week | #372New entry this week | #10New entry this week |
4 WindsurfNew entry this week 15 | #10New entry this week | #10New entry this week | #461New entry this week | #10New entry this week |
5 NeovimNew entry this week 13 | #10New entry this week | #10New entry this week | #552New entry this week | #10New entry this week |
6 ZedNew entry this week 11 | #10New entry this week | #10New entry this week | #644New entry this week | #10New entry this week |
7 Visual StudioNew entry this week 10 | #10New entry this week | #10New entry this week | #738New entry this week | #10New entry this week |
8 GNU EmacsNew entry this week 8 | #10New entry this week | #10New entry this week | #832New entry this week | #10New entry this week |
9 Sublime TextNew entry this week 7 | #10New entry this week | #10New entry this week | #927New entry this week | #10New entry this week |
10 FleetNew entry this week 6 | #10New entry this week | #10New entry this week | #1023New entry this week | #10New entry this week |
The last 12 weeks.
Consensus rank · top 5| Brand | 11 wk ago | 10 wk ago | 9 wk ago | 8 wk ago | 7 wk ago | 6 wk ago | 5 wk ago | 4 wk ago | 3 wk ago | 2 wk ago | 1 wk ago | This week |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Cursor | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | #1 |
| Visual Studio Code | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | #2 |
| JetBrains IDEs (IntelliJ, PyCharm, WebStorm) | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | #3 |
| Windsurf | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | #4 |
| Neovim | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | not ranked | #5 |
The receipts.
Recorded 2026-07-06 · 4 model answersThe unedited answers this ranking is scored from. We record what each model said, word for word. Nothing summarized, nothing editorialized.
ChatGPT7 sources
{"answer":"**My actual buying advice:** choose **Cursor** if you want the strongest AI-first editor today; choose **VS Code + GitHub Copilot** instead if you value a durable, broadly supported default more than having the slickest AI-native experience. AI agents can be wrong, so require review and tests regardless of editor.\n\n1. **Cursor** — The best single recommendation for most developers who want AI to drive real multi-file implementation work, because its Agent mode can explore a codebase, edit across files, run commands, and fix errors while retaining a familiar editor workflow. ([docs.cursor.com](https://docs.cursor.com/agent?utm_source=openai))\n\n2. **Visual Studio Code + GitHub Copilot** — The best conservative long-term choice: it combines the largest mainstream editor ecosystem with local, background, cloud, and third-party agent workflows, configurable instructions/agents, and strong user control. ([code.visualstudio.com](https://code.visualstudio.com/docs/agents/overview?utm_source=openai))\n\n3. **JetBrains IDEs + Junie** — Pick this over Cursor or VS Code when IntelliJ, PyCharm, WebStorm, Rider, GoLand, or similar language-aware tooling is central to your job, since Junie can plan multi-step work, edit code, run tests and terminal commands, and even assist with live debugging in supported IDEs. ([jetbrains.com](https://www.jetbrains.com/help/ai-assistant/junie-agent.html?utm_source=openai))\n\n4. **Devin Desktop (formerly Windsurf)** — A strong choice for developers who want to supervise several local and cloud agents rather than merely chat with one, as it evolved from the Windsurf IDE into an agent command center while retaining the full editor, extensions, keybindings, and terminal workflow. ([cognition.com](https://cognition.com/blog/introducing-devin-desktop?utm_source=openai))\n\n5. **Zed** — The best choice for a fast, modern editor and maximum model/agent flexibility, because it supports its own agent plus external agents and terminal agents, multiple concurrent threads, hosted/API/local models, tool permissions, and MCP. ([zed.dev](https://zed.dev/docs/ai/agents?utm_source=openai))\n\n6. **Visual Studio + GitHub Copilot** — The right answer for serious Windows/.NET/C++ development, where the native IDE and Copilot agent capabilities—including the .NET Upgrade Agent—are more valuable than moving to a VS Code-derived editor. ([docs.github.com](https://docs.github.com/en/copilot/reference/copilot-feature-matrix?utm_source=openai))\n\n7. **Android Studio + Gemini Code Assist** — The best specialized option for native Android teams because Gemini assistance is integrated into Android Studio and supports completions, code generation, chat, and agentic multi-step tasks. ([developers.google.com](https://developers.google.com/gemini-code-assist/docs/overview?hl=en&utm_source=openai))\n\n8. **Xcode + GitHub Copilot** — The practical AI-enabled option for Apple-platform developers who must remain in Xcode, with Copilot support for agent mode, chat, completions, review, MCP, and checkpoints. ([docs.github.com](https://docs.github.com/en/copilot/reference/copilot-feature-matrix?utm_source=openai))\n\n9. **Neovim + GitHub Copilot** — Recommend it only to committed modal-editor users: it preserves an exceptionally efficient keyboard-centric workflow and offers Copilot completions, but lacks Copilot’s full chat and agent-mode experience in Neovim. ([docs.github.com](https://docs.github.com/en/copilot/reference/copilot-feature-matrix?utm_source=openai)
Clauderaw answer
placeholder
Geminiraw answer
Here is a ranked recommendation of the top code editors for developers, from strongest to weakest: 1. Cursor: An AI-first fork of VS Code that natively integrates advanced LLMs for predictive coding and seamless codebase chats. 2. Visual Studio Code: The undisputed industry standard offering unparalleled extensibility and massive community support, including robust integrations for GitHub Copilot. 3. JetBrains IDEs (IntelliJ, PyCharm, WebStorm): The premier choice for heavy, enterprise-grade development with deep language-specific intelligence and powerful built-in tools. 4. Windsurf: A rising AI-native editor by Codeium that excels at autonomous coding agents and comprehensive context awareness. 5. Neovim: A highly customizable, keyboard-centric editor loved by power users for its blazing speed and powerful plugin ecosystem. 6. Zed: A lightning-fast, Rust-based editor that provides built-in AI features along with real-time collaboration without the bloat of Electron. 7. Visual Studio: The heavy-duty powerhouse essential for large-scale .NET, C++, and Windows system application development. 8. GNU Emacs: A legendary, infinitely extensible computing environment for developers who want absolute programmatic control over their unified workflows. 9. Sublime Text: A remarkably snappy, lightweight classic known for its incredible startup speed, smooth performance, and distraction-free interface. 10. Fleet: A promising, next-generation polyglot editor from JetBrains designed to bridge the gap between lightweight text editors and remote IDEs.
Perplexity8 sources
[ { "name": "Cursor", "reason": "**Cursor** is the strongest overall recommendation because it’s a VS Code–style editor with the deepest integrated AI features, excellent multi-file and large-codebase support, and mature agentic tooling that measurably boosts day-to-day developer productivity.[1][2][6]" }, { "name": "VS Code + GitHub Copilot", "reason": "**VS Code with GitHub Copilot** is the safest, most mainstream choice, combining the industry-standard editor with a proven AI assistant that integrates smoothly into existing workflows and is ideal for most developers who don’t want to switch editors.[1][2][6]" }, { "name": "Windsurf", "reason": "**Windsurf** (formerly Codeium Editor) is a VS Code fork focused on AI-first workflows that offers strong agentic capabilities and a compelling free or low-cost option for developers who want an autonomous coding assistant without leaving the VS Code ecosystem.[1][3][4][6]" }, { "name": "JetBrains IDEs + JetBrains AI Assistant (Junie)", "reason": "**JetBrains IDEs with JetBrains AI Assistant** are the top recommendation for Java, Kotlin, and enterprise teams, blending JetBrains’ mature refactoring/debugging tooling with integrated AI that fits naturally into professional, large-team workflows.[3][4][6][7]" }, { "name": "Claude Code", "reason": "**Claude Code** is a standout terminal-first and repo-focused assistant that pairs well with any editor, giving you very strong reasoning on complex codebases, refactors, and automation while letting you keep your existing toolchain.[2][4][6][7][8]" }, { "name": "Playcode (Web AI Editor)", "reason": "**Playcode** is a browser-based AI editor that’s excellent if you want zero setup and fast prototyping, letting AI generate and run code directly in the browser without installing a full desktop IDE.[1]" }, { "name": "Replit + AI", "reason": "**Replit with its AI features** is a strong choice for collaborative and educational coding, offering a cloud IDE where AI assistance, instant hosting, and multiplayer editing make it ideal for learning, rapid experiments, and small web apps.[1]" }, { "name": "Zed + AI integrations", "reason": "**Zed** is a high-performance, minimalist editor with growing AI integrations that is appealing if you prioritize speed and a lightweight experience while still wanting basic AI assistance.[1]" }, { "name": "Neovim + GitHub Copilot / other AI plugins", "reason": "**Neovim with GitHub Copilot or similar plugins** is a good option for power users who live in modal editors and want inline AI completions and chat while keeping their highly customized, keyboard-driven workflow.[2][6]" }, { "name": "Xcode + Xcode AI Assistant", "reason": "**Xcode with Xcode AI Assistant** is a reasonable pick for Apple platform developers who prefer to stay inside the official IDE and want integrated AI help with Swift, UI code, and project scaffolding.[3]" } ]
Cited sources
How this ranking is measured2,100 queries · 4 models · updated this weekLive data
This is the public record for Code Editors & IDEs: the same questions, every model, every week. We ask each model the real questions buyers ask in this category and record which brands it recommends and in what order. A brand recommended at position i scores 100 × 0.85^(i−1) for that model (#1 = 100, #2 = 85, #3 = 72, …); unmentioned brands score 0.
Each model is ranked independently, so the columns disagree when the models disagree. The consensus score is the mean across the 4models recorded this week, and movement compares against last week’s close.
We report what the models say. We don’t editorialize, and brands can’t pay to change their position.
This ranking is live: the numbers come from recorded model answers captured through the official model APIs, scored with the published formula above. The raw answers are on this page under “The receipts.”
- What we measure
- We measure on the official model APIs: the same question, the same settings, the same week, for every brand. Web search is on, so the models can draw on what is live on the web. Nothing is personalized to a user, which is what makes the columns comparable.
- What we don’t
- The consumer apps are a different surface. What a person sees inside a chat app can carry memory, personalization, and live experiments on top of the same model, so its answers can differ from the API’s. We do not measure that surface yet. True browser listings, recorded from the consumer apps, arrive with WDIR Ranked, the Pro product. WDIR Ranked · Coming soon
Updated this week · week of 2026-07-06
Prefer it written out? Read the Best Code Editors & IDEs guide. The same record as an editorial answer, with the reasoning spelled out.
Narrow the record.
The same question, tighter. Each refinement is measured the first time someone opens it.
Branches of this record.
Rankings that share ground with this one. Follow a branch to see who AI recommends next door.
Code Editors & IDEs#1 Cursor · this page
- Cloud Infrastructure Providerssame sector · 10 ranked#1Amazon Web Services (AWS)New entry this week
- App Hosting Platformssame sector · 11 ranked#1RenderNew entry this week
- API Testing Toolssame sector · 14 ranked#1PostmanNew entry this week
- CI/CD Platformssame sector · 14 ranked#1GitHub ActionsNew entry this week
- Backend-as-a-Service Platformssame sector · 10 ranked#1SupabaseNew entry this week