OpenAI deprecation detail

Legacy GPT snapshots on a March 26, 2026 shutdown path

This page is a decision brief for teams still pinned to older GPT snapshot IDs or preview aliases. It keeps the shutdown date, affected names, replacement path, and cutover checks in one source-linked view.

Current state

This migration now has a fixed date and a fixed destination.
Deprecated
OpenAI schedules a March 26, 2026 shutdown for selected legacy GPT snapshots and older preview aliases, with replacement guidance pointing to the gpt-5 or gpt-4.1 family.

Shutdown date

March 26, 2026

Last checked

March 12, 2026

Replacement path

gpt-5 or gpt-4.1 family

Why this matters now

Snapshot drift becomes a shutdown problem once the date is fixed.

Use these points to decide whether the team still has pinned legacy GPT names in production, whether preview aliases hide that risk, and what has to be replaced before the cutoff.

Pinned snapshot IDs will stop serving on March 26, 2026.
OpenAI lists legacy GPT snapshot rows with a fixed shutdown date of March 26, 2026, so teams still pinned to those names are now on a dated migration path.
Older preview aliases are part of the same risk surface.
The deprecations page ties preview aliases like gpt-4-turbo-preview and gpt-4-turbo-preview-completions to the same shutdown group, so alias usage can hide legacy snapshot debt in application config.
The replacement is a current model family, not another dated snapshot.
OpenAI points this shutdown group toward the gpt-5 or gpt-4.1 family, which means teams have to make an explicit replacement choice instead of swapping one frozen snapshot for another.

Official sources

Check the OpenAI pages behind this migration brief.

This page uses a fixed source set so the migration brief stays auditable instead of drifting into secondhand advice.

Deprecations

OpenAI deprecations

Source of record for the March 26, 2026 shutdown date, the affected snapshot IDs and preview aliases, and the official replacement recommendation.

Open official page

Snapshot mapping

The move is from pinned snapshot names to current model families.

These are the legacy names the OpenAI deprecations page places on the March 26, 2026 shutdown clock. The migration work is mostly about finding where those names still exist and replacing them with an actively supported family.

BeforeNowWhat changesSources
gpt-4-0314gpt-5 or gpt-4.1 familyThis dated GPT-4 snapshot is on the shutdown list, so any direct model pin to gpt-4-0314 has to be replaced with a current family choice before the cutoff.
gpt-4-1106-previewgpt-5 or gpt-4.1 familyThis preview snapshot is also on the March 26, 2026 shutdown path, so preview-era configs should be treated as current migration debt rather than as harmless legacy defaults.
gpt-4-0125-previewgpt-5 or gpt-4.1 familyThis later preview snapshot is in the same deprecation group, so applications pinned here still need a direct model-family replacement plan.
gpt-4-turbo-preview and gpt-4-turbo-preview-completionsgpt-5 or gpt-4.1 familyOpenAI ties these preview aliases to the same shutdown path, so alias use in SDK wrappers, config files, or older service defaults still has to be removed explicitly.

Migration checklist

Use the shutdown list to drive a direct replacement sequence.

The order here is meant to surface hidden legacy names first, then force a current-family choice and clear the old IDs from production config before the cutoff date.

Step 1

Inventory every pinned snapshot ID and preview alias.

Search application code, SDK wrappers, infrastructure config, and any model-selection admin tools for the exact legacy names listed on the deprecations page.

Step 2

Choose the replacement family explicitly.

OpenAI points this shutdown group toward the gpt-5 or gpt-4.1 family, so the migration plan should make that choice explicit instead of letting a stale alias keep deciding it implicitly.

Step 3

Remove preview aliases from runtime configuration.

If the application still relies on gpt-4-turbo-preview or other preview-era names, replace those references directly so the shutdown clock is no longer hidden behind aliases.

Step 4

Schedule cutover before March 26, 2026.

Treat the shutdown date as the operational deadline and move production traffic early enough that any fallback, rollback, or benchmark work still happens before the listed removal date.

Verify before cutover

Do not assume a current-family replacement is risk-free just because the API name changes cleanly.

These checks are the minimum confidence gates before any production path should leave the legacy snapshot group behind.

No remaining legacy names in config

Confirm the exact snapshot IDs and preview aliases listed on the deprecations page no longer appear in runtime configuration, environment settings, or admin controls.

Replacement choice is explicit and documented

Verify the team has deliberately selected the gpt-5 or gpt-4.1 family for each affected workload rather than inheriting the replacement through stale defaults or wrapper code.

Tests cover every endpoint that referenced the old snapshot

Re-run the production paths that depended on the legacy names so any pinned-model behavior still hidden in older request paths is found before the shutdown date forces it.

Production traffic moves before the listed date

The shutdown date is the hard line from OpenAI. Treat successful traffic movement before March 26, 2026 as the final completion condition, not just local config edits.

Continue the site

Keep moving through the decision from here.

Use the groups below to move laterally through the decision, not back out into another doc hunt.

Related pages

Stay in the same decision neighborhood instead of backing out to search.

OpenAI deprecations tracker

Provider-specific hub for OpenAI shutdown dates, migration briefs, and replacement paths.

Open page

Assistants API shutdown date

Shutdown-date brief for Assistants API with migration mapping and replacement path.

Open page

GPT-4o preview audio and realtime models

Migration brief for dated preview audio and realtime model names moving to current families.

Open page

Compare pages

Open the pages that turn this topic into a side-by-side decision.

GPT-5.4 vs GPT-5 mini

Side-by-side comparison of GPT-5.4 and GPT-5 mini across price, fit, and tool pressure.

Open page

Cheapest OpenAI model for extraction

Scenario recommendation page for choosing the cheapest workable OpenAI extraction model.

Open page

Replacement pages

Use the likely substitutes, migration targets, or fallback choices as the next click.

GPT-5.4 pricing

Single-model pricing brief for GPT-5.4 across short, long, and batch rows.

Open page

GPT-5 mini pricing

Single-model pricing brief for GPT-5 mini across standard and batch rows.

Open page

GPT-5.4 context and tool support

Limits brief for GPT-5.4 versus GPT-5 mini context windows, output caps, and tool support.

Open page

Source category pages

Trace the source families behind this page instead of opening random docs in isolation.

Deprecation and migration sources

Official shutdown, migration, and replacement references behind lifecycle pages.

Open page

Return

Return to the OpenAI tracker
Go back to the main OpenAI decision surface to compare this snapshot shutdown risk with current cost signals, tool charges, and other lifecycle changes.