🔍 Read the full analysis: A Step-by-Step OpenAI API Migration For GPT-4o, GPT-4.1, And GPT-5-nano on ThorstenMeyerAI.com
Get the latest gadgets delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
OpenAI’s published API schedule lists model retirements on Oct. 23 and Dec. 11, 2026, and Jan. 14, 2027, affecting GPT-4o, GPT-4.1, GPT-5-nano and other models. The source says listed replacements may change costs and behavior, so developers should verify model IDs, pricing and application outputs before switching.
OpenAI’s API deprecation schedule lists GPT-5, GPT-5 mini and GPT-5-nano for retirement on Dec. 11, 2026, followed by listed GPT-4o and GPT-4.1 snapshots on Jan. 14, 2027. A report by Thorsten Meyer, dated Nov. 1 and based on OpenAI’s published schedule, says several other model IDs were already shut down on Oct. 23, leaving developers to check whether their applications still call retired models.
The report identifies three immediate dates: Oct. 23, when a group of older models was retired; Dec. 11, when GPT-5, GPT-5 mini, GPT-5-nano, GPT-5-pro, o3 and o3-pro are listed to leave the API; and Jan. 14, 2027, when listed GPT-4o, GPT-4o mini, GPT-4.1 and GPT-4.1 mini snapshots are scheduled to follow. OpenAI’s deprecations page is the cited source for the dates. The report says calls using already-retired IDs may fail.
For the December group, the report says OpenAI lists GPT-5.6 Sol, GPT-5.6 Terra and GPT-5.6 Luna as replacements for different models. It identifies GPT-6 Luna as the listed replacement for the January GPT-4o and GPT-4.1 snapshots. These are recommendations in the schedule, not evidence that every application will behave identically after changing its model ID.
The source also flags a pricing and behavior change. Its Nov. 1 comparison of standard short-context prices says GPT-6 Luna’s listed input and output rates are lower than GPT-4o’s, while GPT-5-nano’s listed replacement costs more per token. It describes the replacements as reasoning models and warns that hidden reasoning tokens may be billed as output, some sampling parameters may not apply, and low output limits could produce empty responses. Developers should confirm current rates and model-specific documentation before budgeting or deploying.
Infographic 01 · The shutdown calendar
What stops working, and when
OpenAI's API deprecations, as listed on 1 November 2026. GPT-4o and GPT-4.1 already left ChatGPT on 13 February 2026; this round is about apps built on the API.
- 23 Oct 2026Already offAlready shut down
gpt-3.5-turbo,gpt-4,gpt-4-turbo,gpt-4.1-nano,gpt-4o-2024-05-13,o1,o1-pro,o3-mini,o4-mini,gpt-image-1. Calls to these fail today. - 30 Nov 2026Your testsEvals platform, Agent Builder, reusable prompts
Not models, but tools many teams test with. Existing evals became read-only on 31 Oct. Export them first.
- 11 Dec 2026Notified 11 JunGPT-5 family and o3
gpt-5→gpt-5.6-sol·gpt-5-mini→gpt-5.6-terra·gpt-5-nano→gpt-5.6-luna·o3→gpt-5.6-sol·gpt-5-pro,o3-pro→gpt-5.6-sol(pro mode) - 14 Jan 2027Notified 14 OctGPT-4o, GPT-4o mini, GPT-4.1, GPT-4.1 mini
Snapshots
gpt-4o-2024-08-06,gpt-4o-2024-11-20,gpt-4o-mini-2024-07-18,gpt-4.1-2025-04-14,gpt-4.1-mini-2025-04-14→gpt-6-luna. Fine-tuned versions follow their base model. - Jan to Apr 2027Plan aheadAudio, speech, older GPT-5.x
6 Jan:
tts-1and others · 20 Jan: older audio and realtime models · 26 Feb:whisper-1, GPT-4o transcription · 1 Apr:gpt-5.1,gpt-5.4-nano,gpt-5.3-codex
Sources: OpenAI deprecations page (checked 1 Nov 2026) · OpenAI Help Center, model release notes. Dates as listed by OpenAI; shutdowns can be rescheduled.
Powered by Thorsten Meyer AI
Migration Risks for API Applications
The schedule matters to teams because a model ID can be embedded in application code, test fixtures, configuration files or no-code workflows. If a request continues to use a retired ID, an API-dependent feature may stop working. A successful model-name substitution, however, does not by itself establish that the replacement will preserve an app’s formatting, latency, cost or response quality.
Pricing changes can affect operating budgets in either direction. In the source’s comparison, GPT-5-nano to GPT-5.6 Luna rises from $0.05 to $0.20 per million input tokens and from $0.40 to $1.20 per million output tokens. By contrast, its GPT-4o-to-GPT-6-Luna comparison shows lower listed rates. Those figures are tied to the report’s stated pricing check on Nov. 1, 2026, and short-context pricing; they should not be treated as a universal forecast for every workload.
Teams also face a testing constraint: the source says OpenAI’s Evals platform, Agent Builder and reusable prompt objects are scheduled to close on Nov. 30, with existing evals read-only from Oct. 31. If that schedule applies as reported, developers relying on those tools may need to preserve test cases and results elsewhere before the date.
Popular right now
Shark PowerDetect Corded Stick Vacuum with HEPA Filtration, Purple (HZ4001AMZ)
Dyson V8 Cyclone Cordless Vacuum
Vitamix Propel Series 510 Professional-Grade Blender, 3 Programs, Black
As an affiliate, we earn on qualifying purchases.
How the Retirement Schedule Is Set
The report says OpenAI’s deprecations page was checked on Nov. 1, 2026 and that the company emails affected customers. It lists several model IDs as already retired on Oct. 23, including GPT-3.5 Turbo, GPT-4 Turbo, GPT-4.1 nano, the original GPT-4o snapshot and several o-series models. The report distinguishes dated snapshots from short model names and advises treating the commonly used GPT-4o and GPT-4.1 names as affected by the January schedule because their model pages reportedly show deprecated status.
OpenAI’s listed dates are not limited to the models named in the headline. The source describes later 2027 retirements for some speech, audio, transcription and coding models. It also says OpenAI’s policy provides at least six months’ notice for generally available models, subject to exceptions for safety or compliance concerns. The report notes that the Oct. 14 notice for the Jan. 14 retirement is three months, but says the deprecation entry does not explain which notice rule applies.
The article is about the API schedule, not the ChatGPT product. The source says GPT-4o, GPT-4.1 and GPT-4.1 mini had already been removed from ChatGPT on Feb. 13, 2026; that separate change does not establish the status of API access.
“The new Model Spec stops promising a house style.”
— Thorsten Meyer, describing the Model Spec revision dated Oct. 29
Notice Period and Replacement Behavior
The report does not establish why the Jan. 14 retirement notice runs for three months when the policy it cites says generally available models receive at least six months’ notice, except in some cases. OpenAI’s entry, as described in the source, does not say whether an exception applies. The effect of any customer contract or committed-spend arrangement is also not specified; the report says some customers may be able to purchase dedicated capacity for continued access, but gives no eligibility details.
There is also no independent test data in the supplied material showing how the replacement models compare on a particular application’s accuracy, response format, latency or total cost. The Model Spec revision is dated Oct. 29, but its practical effect on individual outputs remains uncertain. The article’s prices and schedule are a snapshot from Nov. 1, 2026, and may change.
Audit Calls and Test Replacements
Developers should first search repositories, deployment settings, test fixtures and automation tools for every model ID in use, including dated snapshot names. They can then compare those IDs with OpenAI’s current deprecations page and account notices, rather than relying solely on a model’s short name. The source recommends checking the latest pricing before revising budgets and asking an OpenAI account team about contract terms or continued-capacity options where relevant.
Before deployment, teams should test listed replacements against representative prompts and application requirements. The report advises making required output formats explicit in prompts instead of relying on defaults, and checking token limits and supported sampling parameters for the selected model. Tests should record response validity, quality and token use so teams can identify behavioral or cost changes before a shutdown date.
The next dates identified in the report are the Nov. 30, 2026 closure of specified evaluation and prompt tools, the Dec. 11 model retirements, and the Jan. 14, 2027 GPT-4o and GPT-4.1 retirements. OpenAI’s live deprecations page and direct customer notices remain the references to check for revisions.
Key Questions
When are GPT-4o and GPT-4.1 API snapshots scheduled to be retired?
The report says OpenAI’s schedule lists affected GPT-4o, GPT-4o mini, GPT-4.1 and GPT-4.1 mini snapshots for Jan. 14, 2027. Check the current deprecations page for model-specific entries or updates.
When are GPT-5, GPT-5 mini and GPT-5-nano scheduled to leave the API?
The source lists Dec. 11, 2026 as the retirement date for GPT-5, GPT-5 mini and GPT-5-nano, alongside several other models.
What models does the report list as replacements?
It says GPT-6 Luna is listed for the January GPT-4o and GPT-4.1 group. For the December group, it identifies GPT-5.6 Sol, Terra and Luna as replacements for different models. Confirm the exact mapping in OpenAI’s current documentation before changing production systems.
Will a replacement model have the same cost and behavior?
Not necessarily. The report’s Nov. 1 price comparisons show differences among model pairs, and it warns that reasoning-token billing, supported parameters and response behavior can differ. Test the replacement with representative requests and verify current rates.
What should developers do first?
Search code, configuration and automation tools for model IDs and dated snapshots, then compare the results with OpenAI’s live deprecation notices. Test proposed replacements before deployment, including required output formats, token limits and cost.
Source: ThorstenMeyerAI.com
