기본 콘텐츠로 건너뛰기

Val Town Is Shutting Down: Move Vals and Crons by Dec 10

In short: Val Town announced on November 10, 2026 that it is winding down. New signups and upgrades stopped that day, code becomes read-only on December 10, 2026, and everything else, including running vals, SQLite and blob storage, stops on January 10, 2027. If you have HTTP endpoints or cron jobs there, the safe plan is to copy the code into a Git repository now, move each cron to a scheduler that runs in UTC the same way, keep your own "last run" checkpoint, and only then point your *.val.run URLs at the new home.

The dates that actually matter

The official wind-down post by Val Town's founder lists four dates. Here is what each one means in practice:

  • November 10, 2026: new signups and plan upgrades closed. Billing was paused the same day and existing subscriptions keep working for free; annual prepaid plans get a partial refund.
  • December 10, 2026: read-only mode. You can no longer edit code, and this is also the last day to add or change val.run redirects. Read-only mode does not include SQLite and blob storage, so data stays reachable a bit longer.
  • January 10, 2027: all other services end. Your vals stop answering HTTP requests and your crons stop firing.
  • February 1, 2027: Val Town plans to open-source the platform.

The practical takeaway: December 10 is your real deadline for anything that needs editing, such as adding an export endpoint for your database or setting redirects. January 10 is when users notice.

Step 1: List what you have and decide what to keep

Val Town's own migration guide suggests starting by deciding which vals are worth moving at all. Many accounts are full of one-off experiments. For each val you keep, write down four things before touching any code:

  1. Trigger type: HTTP, cron, email, or plain script. Vals with a trigger must export a function (the default export when there are several), so the exported function is your entry point.
  2. Dependencies: npm: imports, esm.town imports of other vals, and Val Town standard library helpers such as blob, email or SQLite.
  3. Data: whether it uses its own SQLite database or the older account-global one, and which blob keys it reads and writes.
  4. Public surface: its val.run URL or custom domain, and any webhook provider that calls it.

Val Town also publishes a ready-made prompt for doing this inventory with an AI coding agent through its plugin or MCP server. It asks the agent to list your vals, help pick a platform and propose a plan for you to approve, which is a reasonable way to handle a large account. If you only have a handful of vals, doing it by hand takes less time than setting that up.

Step 2: Get the code into Git

The vt CLI (installed with Deno: deno install -grAf jsr:@valtown/vt) can vt clone a val into a local folder, which is the quickest way to get plain files you can commit. Val Town's docs now recommend the plugin or MCP server over the CLI for most agent work, but for a straight copy, the CLI is enough.

Val Town's GitHub guide adds a few rules worth following even if you migrate by hand:

  • Copy every val your app imports through esm.town into the project and switch to relative imports. Bundling a remote URL does not make the app independent of Val Town, and those URLs will stop resolving.
  • Keep private vals in a private repository. If visibility is unclear, start private.
  • Add a .gitignore before the first commit and leave out env values, .env files, database and blob exports, and logs. Ignoring a file later does not remove it from history.
  • Write a short README with the live URL, how to run and deploy, and the names (not values) of the environment variables.

One more reason not to just move the Deno code as-is: Val Town's migration guide notes that Val Town runs on Deno and links Deno's own announcement about winding down, saying core maintainers will not support Deno beyond October 2027. Val Town recommends moving to Node, Bun, or a provider runtime such as Cloudflare Workers or Vercel Functions.

Step 3: Pick a destination

Val Town's "Where to migrate" page compares providers by built-in features. Cloudflare is the only one ticked for every column (website, SQLite, blob, cron, email, access, AI and functions), and the team says it moved its own blog and docs there. Vercel and Netlify cover website, blob, cron, access and AI but not SQLite or email. Render covers website and cron. The table below focuses on the part that breaks most often in these moves: scheduled jobs.

Destination Where a cron val goes Timezone Gotchas noted in Val Town's guides Built-in SQLite / blob
Cloudflare Worker scheduled() handler + triggers.crons in Wrangler config UTC, like Val Town Last-run time is not passed in; store it yourself. Use Queues or Workflows for long multi-step jobs. D1 / R2
Vercel Vercel Schedules (currently beta) with @vercel/schedules Set per schedule (example uses UTC) Hobby plan has a 15-minute minimum interval and timing jitter; schedules can also exist in preview deployments. No SQLite / Vercel Blob
Render Cron job service in render.yaml UTC Cron jobs are paid services; let failures exit non-zero; no lastRunAt supplied. Neither built in
Netlify Scheduled Function with config.schedule UTC Runs only on published deploys; 30-second scheduled execution limit; next_run is not the last run. No SQLite / Netlify Blobs

For reference, Val Town's own Free plan allowed crons every 15 minutes and Pro every minute, so a Vercel Hobby project matches a former Free-plan cadence but not a one-minute Pro job.

Step 4: Move cron jobs without losing state

This is where I would spend the most care. On Val Town, a cron handler received an Interval object with lastRunAt, and many vals used it to fetch only "what's new since the last run". None of the destinations above hand you that value. If you drop it, the job either reprocesses everything on the first run or silently skips items.

I wrote a small port to check the pattern: a Cloudflare-style scheduled() handler that keeps lastRunAt and a per-item "done" marker in a key-value store (an in-memory stand-in for Workers KV or D1 so it runs locally on Node 22). The core logic:

async function runJob(env, scheduledTime) {
  const last = await env.KV.get("lastRunAt");
  const lastRunAt = last ? new Date(last) : undefined;
  for (const it of items) {
    if (lastRunAt && new Date(it.at) <= lastRunAt) continue;
    if (await env.KV.get("done:" + it.id)) continue; // retry guard
    send(it);
    await env.KV.put("done:" + it.id, "1");
  }
  await env.KV.put("lastRunAt", new Date(scheduledTime).toISOString());
}
export default { async scheduled(event, env) { await runJob(env, event.scheduledTime); } };

Run against two items, then a simulated retry of the same tick, then a new item an hour later, it sent a1, a2 on the first run, nothing new on the retry, and only a3 on the third run. That is the behavior you want: the checkpoint moves forward only after the work is recorded, and a retried run cannot send twice. Val Town's guides make the same point for every destination: record processed IDs so a repeated run does not resend messages or duplicate charges.

Watch the UTC trap

Val Town evaluates cron expressions in UTC, and so do Cloudflare, Render and Netlify. If you copy an expression as-is, the time stays the same. If you rewrite it from memory in local time, it moves. For a job meant to run at 9:00 local time, the UTC expression is 0 0 * * * in Tokyo or Seoul, 0 14 * * * in New York during standard time, and 0 17 * * * in San Francisco during standard time. Also note that a Val Town "simple interval" such as "every two hours" is an elapsed interval, not a wall-clock schedule; the guides warn the two are not always equivalent.

Cut over in the right order

Every destination guide gives the same order: test the handler with disposable data, pause the Val Town cron, enable the new schedule, then watch for one real scheduled run and check its effects. A successful deploy does not prove the job ran. Netlify's guide also warns never to publish a job handler without its schedule or an auth check, because an ordinary function URL would let anyone trigger it.

Step 5: Copy SQLite and blob data

Val Town's Cloudflare guide exports SQLite by adding a temporary HTTP val that streams the tables as SQL, protected by a random bearer token you keep out of the code and URLs. Pause writes while exporting, since streaming pages is not a snapshot, and keep the downloaded file out of Git because it contains plaintext data. Because this needs a new val, do it before December 10, even though storage stays readable until January 10.

For blobs, the guide says to list keys with pagination, copy the actual bytes (not just download URLs), preserve keys and content types, and compare counts, sizes and hashes before switching. A failed or rate-limited listing does not mean the store is empty.

Step 6: Redirect val.run URLs last

Val Town lets you redirect any *.val.run endpoint to its new URL from the val's settings (HTTP section) or from account or org settings under Custom Domains. Redirects keep the path and query string by default. Two details matter: the val stops receiving requests as soon as a redirect is set, so add it only after the new app is verified; and many webhook providers do not follow redirects, so update webhook URLs with each provider directly. You can edit redirects until December 10, 2026, when they become permanent. Custom domains are not covered by this; change their DNS at your domain provider.

Editor's note

Who this is for: anyone with HTTP vals, cron vals or stored data on Val Town, especially small automations (digests, webhooks, bots) that nobody has looked at in months. Less relevant: if you only used Val Town for throwaway scripts, cloning the few you like is enough.

Gotchas I'd check first: lost lastRunAt state (first run reprocesses everything); local-time rewrites of UTC cron expressions; leftover esm.town imports that break after shutdown (search your built output for esm.town and .val.run); Deno-only APIs that do not exist in Workers or Node; and webhooks that ignore redirects.

Common errors and fixes: a migrated cron "doesn't run" on Netlify because the deploy is not published, or on Vercel because the schedule exists only in a preview; a job runs twice after a retry because there is no processed-ID record; a redirect "breaks" a working val because it was set before the new app was live.

FAQ

Will my vals stop working on December 10?

No. December 10, 2026 makes code read-only and locks redirects. Vals keep running until January 10, 2027, and SQLite and blob storage are not part of the December read-only change.

Do I have to keep paying until then?

No. Val Town paused billing on November 10 and keeps existing subscriptions running for free; annual prepaid plans get a partial refund.

Which platform is closest to Val Town?

By Val Town's own feature table, Cloudflare, which has built-in equivalents for websites, SQLite (D1), blob (R2), cron, email and access. Vercel is the next closest if you do not need SQLite or email.

Related on lightmind24: Google Opal Shuts Down Nov 17: Back Up and Move Workflows · AWS Emulator Tutorial: Let Coding Agents Test AWS Locally · Agent Skills: One SKILL.md for Claude Code, Cursor, Copilot

Move the crons first and the redirects last, and you will have a quiet January 10.

Screenshots: see the official wind-down timeline, provider table, cron docs and Cloudflare cron guide, plus my own migration map, at the end of this post. Images: Val Town blog and docs.val.town, and lightmind24 own summary table.

Last verified: 2026-11-12

Sources: Val Town blog: Val Town is winding down (Nov 10, 2026) · Val Town migration guide · Where to migrate · Store code on GitHub · Redirecting val.run URLs · Cloudflare: migrate crons · Vercel: migrate crons · Render: migrate jobs · Netlify: migrate crons · Cloudflare: migrate SQLite · Val Town cron docs · Val Town pricing

lightmind24 is an independent review site, not affiliated with Val Town, Cloudflare, Vercel, Render or Netlify.

Uploaded Image 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...