기본 콘텐츠로 건너뛰기

Windsurf Is Now Devin Desktop: Fixes for Common Problems

Summary: Windsurf, the AI code editor, was renamed Devin Desktop by Cognition on June 2, 2026, and a lot of "Windsurf not working" searches since then come from the rename itself: a new app name, a new terminal command, new docs and a new quota system. This guide walks through the problems Cognition's own troubleshooting pages list, from "stuck on the Free tier" and rate limits to the macOS "damaged" pop-up, Linux launch crashes and a stuck Cascade terminal, with the fix the docs give for each. It closes with a plan table so you can tell a real bug from simply running out of quota.

First, rule out the rename

Cognition's Devin Desktop FAQ says Devin Desktop is "the new name for Windsurf": the same IDE, editor and features, delivered as a normal over-the-air update. Your settings, extensions and plan carry over, and the FAQ states that pricing did not change because of the rename. What did change can look like breakage if you were not expecting it:

  • The app has a new name. It now appears as Devin.app on macOS, Devin.exe on Windows and Devin / devin on Linux. If your company's device management only allows "Windsurf", the renamed app can be flagged or blocked. The FAQ tells IT teams to add Devin to the allowlist.
  • The surf command was replaced. The new terminal command is devin-desktop. Install it from the command palette with "Shell Command: Install 'devin-desktop' command in PATH". The old surf and windsurf commands still ship in ~/.codeium/windsurf/bin/, and the FAQ suggests alias surf="devin-desktop" if your scripts depend on the old name.
  • The Agent Command Center opens first. The multi-agent view is now the default surface. The classic editor is still there; nothing was removed.
  • New data folders. On first launch the app offers "Migrate from Windsurf", which copies settings, keybindings, snippets, workspace state and extensions into new paths such as ~/.config/Devin/ on Linux or %APPDATA%\Devin\ on Windows. Caches, logs and Local History are deliberately skipped. If you chose "Start fresh", nothing is copied, which explains a sudden "where did my settings go" moment.
  • The docs moved. Documentation and the changelog now live on docs.devin.ai. Old docs.windsurf.com links redirect there.

Your rules also keep working. Devin Desktop still reads the legacy .windsurfrules file and .windsurf/rules/, and adds .devin/rules/, which takes precedence when both exist. It also reads AGENTS.md and can import .cursor/rules files.

Account and usage problems

"I paid for Pro, but I'm still on the Free tier"

The official fix is a short sequence: wait a few minutes for the upgrade to sync, then log out of Devin Desktop on the website, restart the IDE, and log back in. Also make sure you are running the latest version. You can confirm your plan by clicking the Devin Desktop widget in the status bar and opening the "Plan Info" tab, or under Settings > Usage & limits.

Rate-limit errors on premium models

The troubleshooting page is candid here: Cognition is subject to rate limits from the model providers it uses and "sometimes hit[s] capacity" on premium models. The advice is to wait a few moments and try again. If the error keeps returning on one frontier model, switching to another model for the task is a practical workaround.

"I hit my limit" is not a bug

In March 2026 the product replaced prompt credits with a quota system. Each plan now has a daily and a weekly allowance that refreshes automatically, measured by how many tokens each request uses, so cost varies by model and by how much context you send. Free models do not count against the quota. When you run out, Free users wait for the next daily or weekly reset; Pro, Teams and Max users can buy extra usage, billed at API list prices. The quota page notes that the daily allowance is more than one seventh of the weekly one, so weekend coders can still use the full week.

To stretch your quota, the docs recommend precise prompts with less unnecessary context, using the lower-cost SWE models for routine tasks, avoiding long sessions when a quick prompt will do, and sticking to one frontier model so caching can cut token use.

Install and launch errors

macOS says "Devin Desktop is damaged and cannot be opened"

Cognition calls this a false positive in macOS security. Try these in order:

  1. Open System Settings > Privacy & Security and click "Allow" or "Open anyway".
  2. Make sure the app lives in /Applications and you launch it from there.
  3. Check that you downloaded the right build: Intel for Intel Macs, Apple Silicon for M-series Macs.
  4. Download the DMG again from the official page, since the check is usually triggered at download time.
  5. With the app and the pop-up closed, run xattr -c "/Applications/Devin.app/".

Windows updates don't appear

If you see "Updates are disabled because you are running the user-scope installation of Devin Desktop as Administrator", the app cannot auto-update while elevated. Start it normally, with user scope, and the update can install.

Linux: the app silently doesn't start or crashes on launch

With the tarball install, this is usually an Electron sandbox permission issue that VS Code shares. The documented fix is to give the bundled chrome-sandbox the right owner and mode:

sudo chown root:root /path/to/windsurf/chrome-sandbox
sudo chmod 4755 /path/to/windsurf/chrome-sandbox

Launching with --no-sandbox also works, but Cognition does not encourage it.

"Devin Desktop failed to start" or a blank Cascade panel

The docs suggest deleting the Cascade data folder, ~/.codeium/windsurf/cascade on Linux and macOS or C:\Users\<YOUR_USERNAME>\.codeium\windsurf\cascade on Windows, then restarting. Be warned: this removes your conversation history and local settings, so treat it as a last step.

Linux: "no space left on device" with plenty of disk space

This one confuses many people. The language server watches workspace files with inotify, and when the kernel's watch or instance limit is used up, Linux returns an ENOSPC error that reads like a full disk. The app opens, but the language server exits right away and AI features stop working. Check your limits with cat /proc/sys/fs/inotify/max_user_watches and cat /proc/sys/fs/inotify/max_user_instances. The docs say typical defaults are 8,192 watches and 128 instances. A temporary fix is:

sudo sysctl fs.inotify.max_user_watches=524288
sudo sysctl fs.inotify.max_user_instances=1024

To make it permanent, append both lines to /etc/sysctl.conf and run sudo sysctl -p, then restart the app. Note the trade-off the docs mention: 524,288 watches can use roughly 512 MB of kernel memory.

Editor and agent quirks

The Cascade terminal looks stuck after a command finishes

Three causes are documented. First, no default terminal profile is set: open Settings, search "terminal default profile", and pick zsh, bash or PowerShell for your OS. Second, a heavy zsh theme such as Oh My Zsh or Powerlevel10k can make the agent think a command is still running; comment out the theme lines in ~/.zshrc to test, and keep a minimal config for the app's terminal if that fixes it. Third, on newer Linux distributions (reported on Fedora 43 and later), systemd "terminal context tracking" emits extra escape sequences that confuse output parsing; the workaround is not sourcing /etc/bashrc from ~/.bashrc, or using a minimal shell config for the app.

Python highlighting is broken or Pylance won't work

Microsoft's Pylance is not the supported path here. Search the extensions panel for "Windsurf Pyright", or paste @id:codeium.windsurfPyright, to install the dedicated extension.

Remote SSH fails with "Undefined error: 0" on macOS

If SSH works from Terminal but fails inside the app, macOS is blocking the app's Local Network access. Go to System Settings > Privacy & Security > Local Network, enable Devin Desktop, and restart. If it isn't listed, try connecting once to trigger the permission prompt; reinstalling re-triggers it if you dismissed it earlier.

Behind a VPN, proxy or firewall

Connection problems on restricted networks usually come down to blocked domains. The docs list the backend domains to allow for all customers, including .codeiumdata.com, update.windsurf.com, .windsurf.com, .codeium.com, .googleapis.com, apis.google.com and static.devin.ai. Separate pages cover proxy settings and TLS inspection tools such as Zscaler.

When nothing works: send diagnostics

Before contacting support, open the agent panel, click the three dots in the top-right corner and choose "Download Diagnostics". The official screenshot of that menu is included at the end of this post.

Plans at a glance

If your "error" is really a quota wall, this is what each individual plan includes, taken from the official pricing page today.

Plan Price Usage Key details
Free $0 Light quota; wait for the daily or weekly reset Limited model availability, no premium models, unlimited Tab completions and inline edits, SWE-2 at slow speed
Pro $20/month Increased quotas plus paid extra usage at API pricing Full model availability, including frontier models; free use of SWE-2; access to Devin Cloud agents
Max $200/month Significantly higher quotas Everything in Pro, aimed at power users
Teams $80/month plus $40/month per full dev seat Each full seat has its own quota Up to 200 users, centralized billing, admin analytics, priority support

Existing subscribers kept their older prices when quotas arrived: the quota page says Pro at $15/month and Teams at $30/seat/month are grandfathered indefinitely. Leftover add-on credits were converted to extra-usage balance at $10 per 250 credits, and that balance can be refunded through support.

Editor's note

Who this is for: anyone who used Windsurf and now sees "Devin" everywhere, and developers on Linux or locked-down work laptops, where most of the documented launch problems live.

Who it's not for: users of the Windsurf JetBrains plugin. The FAQ says that plugin is in maintenance mode and newer JetBrains releases can break it; Cognition recommends moving to Devin over the Agent Client Protocol instead.

Gotchas I noticed while checking the docs: the Cascade reset folder path still says "windsurf", so don't go looking for a "devin" folder there. Free trials of paid plans are not offered, though the Plans and Usage docs say you can request a refund before you use a subscription. And the pricing page's Free column shows no premium models, so a Free user seeing only some models is not a bug.

FAQ

Is Windsurf discontinued?
The name is. Cognition's FAQ says Devin Desktop is the same IDE under a new brand, with settings and plans carried over. Older Windsurf releases can still be downloaded, but Cognition treats every release but the latest as deprecated.

Do I have to use the Devin cloud agent?
No. The FAQ says you can keep using Devin Desktop with local-only agents. Devin Local is the new local agent, a rewrite that inherits your Windsurf settings, and the launch post said the legacy Cascade agent would remain available through July 1 for gradual migration.

Why does my usage run out faster on some days?
Quota is spent per token, so large codebases in context, long sessions and expensive frontier models drain it faster. Free models don't count at all, and the docs say requests to the same model benefit from caching.

Related guides

If you'd rather run an open-source agent inside VS Code with your own API key, see our Cline tutorial on planning, acting and rolling back. Terminal-first coders can compare notes with Aider fixes for edit, token and API key errors. And for a free assistant in the editor you already use, read how to use GitHub Copilot Free in VS Code.

Most "Windsurf is broken" moments after June 2 turn out to be the rename, a quota reset or a Linux permission, so check those three before reinstalling anything.

Image credits: Devin Desktop launch image from the official Devin blog, the "Download Diagnostics" menu image from the Devin Desktop docs, and screenshots of the public devin.ai pricing page captured on 2026-10-30; see them at the end of this post.

Last verified: 2026-10-30

Sources:

Uploaded Image Uploaded Image Uploaded Image Uploaded Image

댓글

이 블로그의 인기 게시물

브루(Vrew) 요금제 정리: 무료 200크레딧으로 어디까지 쓸 수 있나

요약: 영상 자막 편집기 브루(Vrew)는 2026년 4월 22일부터 기능별 사용량 대신 '통합 크레딧' 하나로 요금을 매겨요. 무료 플랜은 매달 200크레딧이 들어오는데, 음성 분석(자동 자막)만 쓰면 최대 200분 분량이라 짧은 영상 몇 편은 충분하지만, 영상 초반의 Vrew 마크를 지우려면 유료 플랜이 필요해요. 이 글은 브루 공식 가격 페이지와 고객센터 공지를 2026년 10월 23일에 다시 확인해, 크레딧 계산법과 플랜 고르는 기준을 정리했어요. 목차 브루 요금 구조: 통합 크레딧이란 1크레딧으로 할 수 있는 일 무료 플랜으로 되는 것과 안 되는 것 유료 플랜 4종 비교표 월 구독·연 구독·기간 이용권, 무엇이 이득일까 상황별 추천: 학생·직장인·유튜버 크레딧이 모자랄 때: 사용 기록 지우개 업그레이드·환불 전에 알아둘 규칙 운영자 메모, 자주 묻는 질문 1. 브루 요금 구조: 통합 크레딧이란 예전 브루는 플랜마다 음성 분석 몇 분, AI 목소리 몇 자, 번역 몇 자처럼 기능별 한도를 따로 줬어요. 그런데 보이저엑스(브루 운영사)가 실제 이용 데이터를 보니 한두 가지 기능만 집중해서 쓰는 사람이 훨씬 많았다고 해요. 그래서 2026년 4월 22일 업데이트부터는 플랜마다 정해진 크레딧을 받고, 어떤 기능에 쓸지는 내가 정하는 방식으로 바뀌었어요. 플랜별 매달 지급량은 Free 200 · Light 2,000 · Standard 10,000 · Business 12,000크레딧 이에요. 공식 공지에 따르면 무료 사용자는 매달 1일에, 유료 사용자는 매월 결제일에 충전되고, 남은 크레딧은 다음 달로 넘어가지 않아요. 바뀐 뒤 무엇이 좋아졌나 공지에 나온 예시가 이해하기 쉬워요. Standard 플랜에서 AI 목소리만 쓰면 예전에는 최대 50만 자였지만, 지금은 1만 크레딧을 모두 AI 목소리에 몰아 쓸 수 있어 최대 100만 자가 돼요. 반대로 여러 기능을 골고루 쓰는 사람은 크레딧이 기능끼리 ...

에이닷 전화 vs 익시오: 통화 녹음·AI 요약, 내 통신사·폰에 맞는 앱 고르기

요약: 통화 녹음과 AI 요약을 쓰고 싶다면 먼저 내 통신사와 휴대폰 종류부터 확인해야 해요. LG유플러스 고객은 무료 앱 익시오, SKT 고객은 에이닷 전화가 가장 기능이 많고, 다른 통신사의 아이폰 사용자는 iOS에 들어 있는 기본 통화 녹음·전사 기능이 현실적인 선택이에요. 이 글은 2026-10-30에 SKT·LG유플러스·Apple 공식 페이지와 도움말을 직접 확인해 세 가지를 비교했어요. 왜 '앱 기능'보다 '내 조건'을 먼저 봐야 할까요 통화 요약 앱 비교 글을 보면 기능 목록부터 나열하는 경우가 많아요. 그런데 실제로 써 보려는 사람에게 가장 먼저 걸리는 건 기능이 아니라 이용 조건 이에요. 익시오는 LG유플러스 고객 전용이고, 에이닷 전화는 누구나 설치할 수 있지만 통역콜·baro 같은 기능은 SKT 고객에게만 열려 있어요. 아이폰은 운영체제 구조상 앱이 통화를 마음대로 녹음하기 어려워서, 통신사 앱이 안 되면 iOS 기본 기능에 기대야 해요. 그래서 이 글은 "어떤 앱이 제일 좋은가"보다 "내 통신사·내 폰에서 무엇을 쓸 수 있나" 순서로 정리했어요. 기능 차이는 그다음이에요. 한눈에 비교: 에이닷 전화 vs 익시오 vs 아이폰 기본 통화 녹음 구분 에이닷 전화 (SKT) 익시오 ixi-O (LG유플러스) 아이폰 기본 통화 녹음 (Apple) 누가 쓸 수 있나 통신사 구분 없이 앱 설치 가능. 단, 통역콜·baro·그룹통화 발신 등은 SKT 고객 전용 U+ 모바일 고객. 타 통신사, 만 14세 미만, 선불요금제 고객은 제한 (공식 사이트 FAQ상 알뜰폰 중 KB리브엠은 예외) 지원 국가·언어의 iPhone 사용자. 1명과 통화할 때만 녹음 가능 지원 OS·기기 Android 10 이상, iOS 16 이상 iOS 17 이상 iPhone 12 이상 / Android...

타입캐스트 vs 클로바더빙: 수익 나는 채널에 쓸 때 무료·유료 조건 비교

요약 타입캐스트와 클로바더빙은 둘 다 글만 입력하면 한국어 AI 목소리를 만들어 주는 서비스지만, 무료로 쓸 수 있는 범위가 꽤 다릅니다. 클로바더빙 무료는 매달 15,000자까지 쓸 수 있는 대신 수익이 생기는 채널에는 쓸 수 없고, 타입캐스트 무료는 처음에 주는 약 5분 분량의 다운로드로 끝나는 체험판에 가깝습니다. 광고 수익이 붙은 유튜브나 쇼핑몰 영상에 넣을 목소리라면 처음부터 유료 플랜 조건을 보고 고르는 편이 안전합니다. 왜 이 두 서비스를 같이 비교하나 쇼츠 내레이션, 수업 자료 더빙, 상품 소개 영상처럼 "내 목소리 대신 AI 목소리"가 필요할 때 한국 사용자들이 가장 많이 찾는 곳이 네오사피엔스의 타입캐스트 와 네이버의 클로바더빙 입니다. 기능 소개만 보면 비슷해 보이지만, 실제로 문제가 되는 건 음질보다 "이 목소리를 내 채널에 써도 되나" 입니다. 그래서 이 글은 기능 나열보다 무료 사용 범위, 출처 표기 의무, 유료 플랜 가격과 다운로드 한도를 중심으로 정리했습니다. 모든 수치는 2026년 10월 25일에 두 서비스의 공식 요금 페이지와 고객센터 도움말에서 직접 다시 확인한 내용입니다. 한눈에 보는 비교표 항목 타입캐스트 무료 타입캐스트 베이직 클로바더빙 Free 클로바더빙 Standard 월 요금 0원 9,900원 (연간 결제 시 10% 할인) 0원 19,900원 (네이버 클라우드 플랫폼에서 결제) 무료·포함 사용량 음성 생성·재생 무제한, 다운로드는 최초 3,000크레딧(약 5분) 1회 지급 월 30,000크레딧(약 60분) 월 15,000자, 다운로드 월 20회, 프로젝트 누적 5개 (매월 1일 갱신) 월 30,000자, 다운로드 월 40회, 프로젝트 누적 20개 쓸 수 있...