Vercel Hobby Deployment Retention Changed: Protect Your Rollback Path
Vercel changed Hobby deployment retention on September 16, 2026. Past the 30-day retention window, each Hobby project now keeps its three most recent production deployments plus its three most recent deployments of any type, while preview deployments no longer receive their own separate protection. Vercel says the current production deployment, aliased deployments and active-branch deployments remain protected, and Hobby teams have 10 GB of Deployment Storage. Treat hosted deployment history as a convenience layer rather than your only recovery plan: keep source control, database backups or migrations, environment/config documentation and a tested redeploy path outside old preview URLs.
A Vercel Hobby project has fewer old preview deployments than you expected
You rely on an old deployment URL as your only rollback or QA reference
Deployment Storage is approaching the Hobby plan's 10 GB allocation
An AI-built website has many frequent previews and no documented recovery procedure
Possible causes
Vercel changed the protected deployment set for Hobby projects in September 2026
Preview deployments no longer receive a separate protected allocation after the retention window
Frequent builds can accumulate deployment storage even when the application repository itself is small
Teams sometimes confuse immutable hosted deployments with source control, database backup and disaster recovery
HOW TO FIX IT
Work from the safest step to the harder repair.
Step 1. Open the project's Deployments and Usage views and identify which production, preview and branch deployments are currently important. Do not assume a historical preview will remain available indefinitely simply because its URL still works today.
Step 2. Understand the September 16 Hobby rule before pruning anything: Vercel says each project now keeps the three most recent production deployments plus the three most recent deployments of any type past the 30-day retention window. The current production deployment, aliased deployments and active-branch deployments have additional protection described by Vercel.
Step 3. Keep the application source in a version-control system you control. Record the exact commit or release tag associated with important production deployments so an old hosted artifact is not the only record of working code.
Step 4. Back up state separately. A deployment snapshot is not a database backup, object-storage backup, payment record, DNS backup or secret manager. Preserve migrations and restore procedures for data the application needs to function.
Step 5. Document environment and build configuration without copying secrets into a repository or public note. Record variable names, runtime versions, build commands and external-service dependencies so a clean redeploy can be reconstructed safely.
Step 6. Before a risky production change, create an explicit rollback plan: identify the last known-good deployment or commit, verify that its database assumptions remain compatible, and know whether rollback requires traffic promotion, a redeploy or a data migration reversal.
Step 7. If Deployment Storage approaches the plan limit, inspect large static assets, generated artifacts and repeated builds before deleting recovery evidence. Vercel states that going over the Hobby storage limit can block deployments until storage is freed.
Step 8. If a needed deployment was removed by retention, check Vercel's Recently Deleted workflow promptly. Vercel documents a 30-day restoration window for recently deleted deployments, but do not make that window your long-term backup strategy.
Step 9. For AI-assisted development, make checkpointing explicit. Agent-generated changes should land in understandable commits or release points with tests and notes rather than depending on a long list of ephemeral previews as the only history.
Step 10. After changing retention or cleanup practices, test one complete recovery drill in a non-destructive environment: rebuild from source, restore required configuration, run typecheck/tests, and verify a representative user journey.
Need the actual code? Go to GenesisCodeDoctor.com to search the Code Store or request code for the exact platform, error, and repair you are working on.
Tag or commit known-good releases before high-risk changes and keep migrations reversible where the data model permits it.
Use separate, access-controlled backups for databases and files; deployment retention is not a substitute for data backup.
Before restoring or promoting an older build, verify authentication, payment/webhook, database-schema and API compatibility so rollback does not create a second incident.
STOP AND GET HELP WHEN
Do not turn a repair into a larger outage.
Do not delete the only known-good deployment while a production incident is unresolved.
Do not store secrets in Git, public deployment notes or screenshots merely to make a rebuild easier.
Do not assume rolling frontend code backward is safe when a newer database migration or external API change is incompatible.
Do not promise that a hosted preview or Recently Deleted item will remain recoverable beyond the provider's current documented policy.
HOW GENESIS HANDLES IT
Diagnose the exact failure before choosing a repair.
Genesis separates the visible symptom from the underlying technical cause. Run the supported diagnostic first, review the evidence, and then use a matching repair only when the failure is actually verified.
Platform scope
Vercel Hobby projects, AI-built sites, Next.js apps and small teams using Vercel preview/production deployments
Category
Deployment · Vercel · Production readiness
Last updated
2026-09-22
REPAIR PROFILE
Know the complexity before you edit.
Difficulty
Beginner
Change risk
Medium
These labels describe implementation complexity and blast radius, not a guaranteed repair time.
AUTHORITATIVE SOURCES
Verify time-sensitive platform details at the source.
How many old deployments does Vercel Hobby now protect?
Vercel's September 16, 2026 announcement says each Hobby project keeps its three most recent production deployments plus its three most recent deployments of any type past the 30-day retention window, with additional protections for the current production, aliased and active-branch deployments.
Does Vercel deployment retention replace Git backups?
No. Deployment history is provider-hosted artifact history. Source control, database backups, environment documentation and a tested rebuild path are separate recovery controls.
Can old Vercel deployments be restored after deletion?
Vercel documents a Recently Deleted workflow that can restore eligible deleted deployments within 30 days. Treat that as a short recovery window, not permanent archival storage.
Can Deployment Storage prevent new deployments on Hobby?
Yes. Vercel says Hobby teams have 10 GB of Deployment Storage and that exceeding the limit can block deployments until enough storage is freed.
RELATED REPAIR GUIDES
Continue from easy to harder.
NEED THE CODE?
Get the code you need from Genesis Code Doctor.
Once you know the platform and the verified problem, search the Genesis Code Doctor Code Store for a matching package. If the exact integration or repair is not there, use Request a Code and describe the platform, official documentation, desired behavior, and sanitized error—never send your secret key.
Start with a free diagnostic. If Genesis verifies a problem and a compatible treatment exists, continue to the matching Code Store product or repair path. If you cannot find the exact code you need, request it at GenesisCodeDoctor.com rather than forcing a generic snippet into the wrong platform.
GENESISCODE DOCTOR AI