🔍 Read the full analysis: When GPT-4o, GPT-4.1, And GPT-5-nano Leave The API: Breakage And Costs on ThorstenMeyerAI.com
Get hardware and tech essentials 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 retirement schedule includes model shutdowns on Dec. 11, 2026, and Jan. 14, 2027; several older model IDs were already retired on Oct. 23. The listed replacements can change token costs and response behavior, so developers need to check model IDs, pricing and app-specific tests before switching.
OpenAI has set Dec. 11, 2026, for retiring GPT-5, GPT-5 mini, GPT-5 nano, GPT-5 Pro, o3 and o3 Pro from its API, followed by listed GPT-4o and GPT-4.1 snapshots on Jan. 14, 2027. The schedule matters to developers because calls to retired model IDs may fail, while the recommended replacements can have different prices and response behavior.
According to the deprecations page reviewed by ThorstenMeyerAI.com on Nov. 1, 2026, a group of older models was already shut down on Oct. 23. The list includes gpt-3.5-turbo, gpt-4-turbo, gpt-4.1-nano, the original gpt-4o-2024-05-13 snapshot, o1, o3-mini, o4-mini and gpt-image-1. Any application still calling one of those IDs may already be receiving failed requests.
For the Dec. 11 retirements, OpenAI’s listed mappings include GPT-5 to gpt-5.6-sol, GPT-5 mini to gpt-5.6-terra, and GPT-5 nano to gpt-5.6-luna. The Jan. 14 list covers dated GPT-4o, GPT-4o mini, GPT-4.1 and GPT-4.1 mini snapshots, with gpt-6-luna listed as their recommended replacement. The source says fine-tuned versions follow their base model’s shutdown schedule and that the listed replacements do not support fine-tuning.
The price comparisons in the source use standard short-context rates per million tokens, checked on Nov. 1. GPT-4o is listed at $2.50 per million input tokens and $10 per million output tokens, compared with $0.10 and $0.50 for gpt-6-luna. The direction differs for some GPT-5 users: GPT-5 nano’s listed replacement costs $0.20 input and $1.20 output, compared with $0.05 and $0.40 for GPT-5 nano. GPT-5 mini’s replacement is listed at $2 input and $12 output, versus $0.25 and $2 for GPT-5 mini.
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
A model swap is not necessarily a drop-in change. The source describes the listed replacements as reasoning models, where hidden reasoning tokens count as output for billing. That can alter total usage costs even when an application sends the same visible prompt. It also reports that some sampling parameters need to be removed and that a tight output limit can result in an empty answer that is still billed.
These differences affect software teams, businesses and users whose tools rely on stable automated responses. A customer-support workflow, internal assistant or no-code automation might break if it continues to call a retired ID; it might also continue running but return differently structured or more expensive responses after migration. The article’s pricing figures are a snapshot, not a guarantee of future charges, and actual spending depends on usage.
There is a separate behavior consideration: OpenAI’s Model Spec was revised on Oct. 29, according to the source. The revision removes much of the detailed style guidance, including a default preference for Markdown, and says models should follow the intent of instructions. Developers relying on a specific format should state that format explicitly and test whether the replacement follows it. The source also says OpenAI acknowledges that its models “do not yet fully reflect” the Spec.
Products Worth Considering
Apple MacBook Pro 16-inch (2021) with M1 Max, 64 GB RAM, 2 TB SSD – Space Grey
Apple MacBook Pro 14 2021 with M1 Max, 64GB RAM, 2TB SSD, Silver
Apple MacBook Pro 16.2-inch with M5 Pro, 18-core CPU, 20-core GPU, 24GB Memory, 1TB SSD, Space Black
As an affiliate, we earn on qualifying purchases.
Two Retirement Dates, One Earlier Cutoff
The schedule distinguishes between dated model snapshots and familiar short names such as “gpt-4o.” The source says the listed snapshots underpin those commonly used names and that the model pages for gpt-4o, gpt-4o-mini, gpt-4.1 and gpt-4.1-mini already show a deprecated status. It advises treating those short names as part of the Jan. 14 retirement.
The API changes are distinct from ChatGPT product changes. ThorstenMeyerAI.com reports that OpenAI removed GPT-4o, GPT-4.1 and GPT-4.1 mini from ChatGPT on Feb. 13, 2026; the announced dates in this schedule concern API access, and therefore applications built on the API. The source also lists a Nov. 30 shutdown for OpenAI’s Evals platform, Agent Builder and reusable prompt objects, which may affect teams preparing or testing migrations.
OpenAI’s deprecations page is described as promising at least six months’ notice for generally available models, except when safety or compliance concerns require a faster schedule. The source notes that the Jan. 14 retirement notice was dated Oct. 14, giving three months, but says the notice does not identify which exception, if any, applies.
Open Questions for API Customers
The schedule and prices cited here were checked by ThorstenMeyerAI.com on Nov. 1, 2026; they may change. The source does not include a response from OpenAI explaining why the Jan. 14 group has three months’ notice despite the stated six-month policy for generally available models. It says customers with contracts or committed spend may want to ask their account teams about the timeline.
It is also not clear from the source how each customer’s actual costs will change after migration. Listed token rates do not account for every workload, prompt length, caching pattern or use of reasoning tokens. Nor does the material establish how every application will behave on replacement models: developers will need to test their own prompts, formatting requirements and output limits.
The source says OpenAI may allow some customers to buy dedicated capacity for continued access after a shutdown, but does not identify who qualifies, what the terms would be or which retired models are covered. Those details remain to be confirmed with OpenAI.
Check IDs Before the Deadlines
Teams should first search code, configuration files, test fixtures and no-code workflows for every model ID they use, then distinguish IDs already retired from those scheduled for December and January. The source recommends testing replacement models before cutoffs, including checks for task accuracy, output format, latency, token use and cost. It also advises revisiting output limits and removing sampling parameters that the replacements do not support.
Customers with affected contracts or committed spend can ask OpenAI account teams about the Jan. 14 notice period and possible dedicated-capacity arrangements. Developers should recheck the official deprecations and pricing pages before budgeting or making a production change, since the cited tables reflect the status on Nov. 1. The next scheduled milestone in the source is the Nov. 30 shutdown of Evals, Agent Builder and reusable prompt objects, followed by the Dec. 11 GPT-5 and o3 retirements.
Key Questions
When do GPT-5 nano and the other listed GPT-5 models leave the API?
The source lists Dec. 11, 2026, for GPT-5, GPT-5 mini, GPT-5 nano, GPT-5 Pro, o3 and o3 Pro.
When are the listed GPT-4o and GPT-4.1 snapshots scheduled to retire?
The listed GPT-4o, GPT-4o mini, GPT-4.1 and GPT-4.1 mini snapshots are scheduled for Jan. 14, 2027. The source says to treat the familiar short names as part of that date as well.
Have any of the models already stopped working?
Yes. The source says several IDs, including gpt-4.1-nano, gpt-4-turbo, gpt-3.5-turbo and the original GPT-4o snapshot, were shut down on Oct. 23, 2026. Requests using retired IDs may fail.
Will migration necessarily make API use cheaper?
No. The listed rates vary by model. The source’s comparisons show lower listed rates for some GPT-4o and GPT-4.1 replacements, but higher rates for the recommended GPT-5 nano and GPT-5 mini replacements. Reasoning tokens are billed as output, and actual costs depend on usage.
What should developers test before switching?
Check whether replacement models handle the application’s tasks, formatting instructions, output limits and supported parameters as required. Compare token usage and costs as well as answer quality, and recheck OpenAI’s current deprecation and pricing pages before the migration.
Source: ThorstenMeyerAI.com
