🔍 Read the full analysis: When To Migrate From GPT-4o, GPT-4.1, And GPT-5-nano In The OpenAI API 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 API deprecation schedule, as reported by ThorstenMeyerAI.com on Nov. 1, 2026, sets shutdowns for GPT-5, GPT-5 mini and GPT-5-nano on Dec. 11, followed by GPT-4o and GPT-4.1 snapshots on Jan. 14, 2027. The listed replacements can change both token costs and model behavior; some already-deprecated IDs stopped working on Oct. 23. The dates, prices and migration details here are based on the source article’s account of OpenAI documentation and should be checked against the current API notices.
OpenAI plans to remove GPT-5, GPT-5 mini and GPT-5-nano from its API on Dec. 11, 2026, then shut down listed GPT-4o and GPT-4.1 snapshots on Jan. 14, 2027, according to a Nov. 1 report by ThorstenMeyerAI.com citing the company’s deprecation page. The changes affect developers running applications and automations through the API, who will need to move to listed replacements and check for differences in price, request settings and output behavior.
The report says the Dec. 11 schedule covers gpt-5, gpt-5-mini, gpt-5-nano, gpt-5-pro, o3 and o3-pro. Its listed replacements are in the GPT-5.6 family: gpt-5 maps to gpt-5.6-sol, gpt-5-mini to gpt-5.6-terra, and gpt-5-nano to gpt-5.6-luna. The report says o3 maps to gpt-5.6-sol, while gpt-5-pro and o3-pro map to that model with its pro reasoning mode.
On Jan. 14, the listed GPT-4o and GPT-4.1 dated snapshots are scheduled to leave the API. The report identifies gpt-4o-2024-08-06, gpt-4o-2024-11-20, gpt-4o-mini-2024-07-18, gpt-4.1-2025-04-14 and gpt-4.1-mini-2025-04-14, with gpt-6-luna listed as the replacement for all of them. It warns that developers should not assume short model names are exempt: the article says their model pages are marked deprecated and advises treating them as part of the January deadline.
Some IDs have already passed their reported shutdown date. The source says models including gpt-4.1-nano, gpt-4-turbo, gpt-3.5-turbo, o1 and o4-mini were shut down Oct. 23, alongside the original GPT-4o snapshot. The same report notes a separate Nov. 30 closure of OpenAI’s Evals platform, Agent Builder and reusable prompt objects, with existing evals read-only since Oct. 31.
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 Can Change Cost and Output
Replacing a model ID is not necessarily a like-for-like change. The source report characterizes the listed successors as reasoning models, which can use hidden reasoning tokens billed as output. It says developers may need to remove or change certain sampling parameters, and that a low output-token cap can result in a billed response with no visible answer. These are operational issues to check in actual application requests, not just in a configuration diff.
The price impact also varies by model. Using short-context prices reported from OpenAI’s pricing page, the article puts GPT-4o at $2.50 per million input tokens and $10 per million output tokens, compared with $0.10 and $0.50 for gpt-6-luna. In contrast, it reports GPT-5-nano at $0.05 input and $0.40 output, versus $0.20 and $1.20 for gpt-5.6-luna. The report says GPT-5 mini’s listed replacement, gpt-5.6-terra, is also more expensive per token. These figures are a dated snapshot, not a forecast of each application’s bill; actual usage and current pricing need checking.
For teams, the stakes include service continuity, budgets and user-facing behavior. A model that still answers may format data differently or handle an instruction in another way. Applications that rely on strict JSON, a particular tone, or a fixed response length need regression tests before migration.
Products Worth Considering
Apple MacBook Air 13-inch with M5 chip, 16GB memory, 1TB SSD, and 12MP Center Stage camera (Midnight)
Apple MacBook Air 15-inch (2026) with M5 chip, 16GB memory, 512GB SSD, Silver
Logitech MX Brio Ultra HD 4K Webcam, Graphite
As an affiliate, we earn on qualifying purchases.
API Dates Differ From ChatGPT
This schedule concerns the OpenAI API, which developers use to power their own applications, bots and workflows. It is distinct from ChatGPT: the source says GPT-4o, GPT-4.1 and GPT-4.1 mini were removed from ChatGPT on Feb. 13, 2026. That earlier change does not itself establish when an API model stops working.
ThorstenMeyerAI.com says it checked OpenAI’s deprecation page on Nov. 1, 2026, and that the company says it emails affected customers and documents deprecations there. The source also reports an Oct. 29 revision to OpenAI’s Model Spec, which it describes as removing much of the detailed style guidance, including a default preference for Markdown. It quotes OpenAI as saying its models “do not yet fully reflect” the Spec. For developers, the practical point is to specify required formats and instructions in their API prompts rather than relying on an assumed house style.
The report says fine-tuned versions follow the shutdown schedule of their base models and that the listed replacements do not support fine-tuning. That could constrain migrations for teams whose applications depend on customized models, making it necessary to confirm available options directly with OpenAI.
“The new Model Spec stops promising a house style.”
— ThorstenMeyerAI.com, describing the revised Model Spec
Confirm Account-Specific Migration Terms
The report’s schedule and price comparisons reflect the information it says was available on Nov. 1, 2026; both model availability and prices can change. The article does not reproduce OpenAI’s full notices, so developers should verify exact model IDs, dates and replacement availability against the current deprecation and pricing pages before acting.
The report says OpenAI’s policy promises at least six months’ notice for generally available models, subject to faster action for safety or compliance reasons. It observes that the notice for the Jan. 14 GPT-4o and GPT-4.1 shutdowns runs from Oct. 14, or about three months, and says the notice entry does not explain which rule applies. Whether an exception applies, and whether any specific customer has contractual or dedicated-capacity options, remains unclear from the material provided.
It is also not established that every application will see the same results after migration. The source does not provide independent benchmark results, application-specific testing, or a full account of how the replacements compare across tasks. Its token prices cover short-context rates, not every possible usage pattern.
Test Before Each Shutdown Date
Teams using affected models should first inventory model IDs across code, deployment settings, test fixtures and no-code workflows. Then they can test the listed replacement against representative real requests, comparing answer quality, formatting, latency, token use and cost. Requests that require a specific format should state it explicitly, and teams should check parameters and output limits against the reasoning-model behavior described in the report.
The immediate dates to track are Nov. 30 for the reported Evals, Agent Builder and reusable prompt-object changes; Dec. 11 for the GPT-5 and o3-family retirements; and Jan. 14, 2027 for the listed GPT-4o and GPT-4.1 snapshots. Developers with fine-tuned models, dedicated-capacity arrangements or committed spending should ask OpenAI how those arrangements are affected. The source says OpenAI may offer some customers dedicated capacity after shutdown, but it does not specify eligibility or terms.
Because model schedules and prices are subject to change, the next step is to recheck OpenAI’s notices and pricing close to deployment, then roll out the migration with monitoring and a rollback plan where available. The report does not say whether OpenAI will revise the listed deadlines or replacement mappings.
Key Questions
When are GPT-5-nano and GPT-5 mini scheduled to leave the API?
The source report says Dec. 11, 2026. It lists gpt-5.6-luna as the replacement for GPT-5-nano and gpt-5.6-terra for GPT-5 mini. Confirm current dates and model availability with OpenAI before migration.
When are GPT-4o and GPT-4.1 API snapshots scheduled to shut down?
The report lists Jan. 14, 2027 for specified GPT-4o, GPT-4o mini, GPT-4.1 and GPT-4.1 mini snapshots. It says gpt-6-luna is the listed replacement and advises treating the short model names as part of the same deadline.
Will switching to a replacement model reduce API costs?
Not in every case, according to the price comparison in the Nov. 1 report. Its listed GPT-4o replacement is cheaper per token, while the reported GPT-5-nano and GPT-5 mini replacements cost more per token. Prices depend on model, input and output usage, and context length; recheck the current pricing page.
What should developers test before switching?
Test representative requests for response quality, required formatting, token use, costs and output limits. The source warns that reasoning tokens are billed as output and that some sampling parameters may need to change. Check any fine-tuning dependency and verify exact model IDs and shutdown dates with OpenAI.
Source: ThorstenMeyerAI.com
