🔍 Read the full analysis: The OpenAI API Changes Coming 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 schedule retires GPT-5, GPT-5 mini and GPT-5-nano from the API on December 11, 2026, followed by specified GPT-4o and GPT-4.1 snapshots on January 14, 2027. The listed replacements differ in price and behavior, while some older model IDs have already been shut down. Developers should check exact IDs and test migrations; the source says OpenAI’s schedule and pricing were checked November 1, 2026.
OpenAI has set API shutdown dates for specified GPT-5 models on December 11, 2026, and GPT-4o and GPT-4.1 snapshots on January 14, 2027, requiring developers using those model IDs to plan a migration. OpenAI’s deprecations page lists newer models as replacements, but the change can affect both API availability and operating costs; some listed models were already switched off on October 23.
According to OpenAI’s deprecations page, the December 11 shutdown covers gpt-5, gpt-5-mini and gpt-5-nano, as well as gpt-5-pro, o3 and o3-pro. The listed replacements are gpt-5.6-sol for GPT-5 and o3, gpt-5.6-terra for GPT-5 mini, and gpt-5.6-luna for GPT-5-nano. OpenAI lists gpt-5.6-sol with its pro reasoning mode for users of GPT-5-pro and o3-pro.
On January 14, 2027, the schedule covers the dated snapshots 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. OpenAI lists gpt-6-luna as the replacement for those models. The source notes that the plain-name model pages already show “Deprecated,” and advises treating the short names as part of the January schedule. Developers need to check the exact model ID in their code, not just the label they use informally.
Several older IDs were scheduled to stop working on October 23, 2026, including gpt-4.1-nano, the original GPT-4o snapshot, gpt-4-turbo, gpt-3.5-turbo, o1 and o4-mini. The source reports that calls using those IDs may already fail. It also says OpenAI’s schedule includes fine-tuned models, which follow their base model’s shutdown date, and that the replacements listed do not support fine-tuning.
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 Costs and Behavior Changes
The deadlines matter because a model name in a production configuration, test fixture or automated workflow can determine whether an API request continues to work. A migration is not just a matter of replacing an identifier: teams may need to review prices, output quality, latency and request settings, then test whether the new model fits the application’s use.
The prices reported by the source show why the impact will vary. At the listed short-context rates per million tokens, GPT-4o is priced at $2.50 for input and $10 for output, compared with $0.10 and $0.50 for gpt-6-luna. GPT-4.1 is listed at $2 input and $8 output, while GPT-4.1 mini is $0.40 and $1.60; both are mapped to the same replacement. Those figures suggest lower listed rates for those comparisons, but actual spending depends on usage and the applicable pricing tier.
The direction differs for smaller GPT-5 models. The source lists GPT-5-nano at $0.05 input and $0.40 output, and gpt-5.6-luna at $0.20 and $1.20. GPT-5 mini is listed at $0.25 and $2, compared with $2 and $12 for gpt-5.6-terra. These are the source’s prices for prompts up to 272,000 tokens, checked November 1, 2026; they should not be treated as a universal bill estimate.
Behavior is another migration risk. The replacements are described in the source as reasoning models. Hidden reasoning tokens are billed as output, some sampling parameters must be removed, and a tight output-token limit can result in an empty answer that still incurs a charge. Applications built around a particular response format or latency profile should be tested rather than assumed to behave identically.
Products Worth Considering
Apple MacBook Air 15-inch (2026) with M5 chip, 16GB memory, 512GB SSD, Silver
DJI Osmo Mobile 8 Standard Combo Gimbal Stabilizer for Phone
Bose QuietComfort True Wireless Noise-Cancelling Earbuds, Black
As an affiliate, we earn on qualifying purchases.
The API Timeline and Model Spec
This is an API change, not a new ChatGPT retirement notice. The source says OpenAI removed GPT-4o, GPT-4.1 and GPT-4.1 mini from ChatGPT on February 13, 2026; the dates in the current schedule concern applications and services calling the API. The distinction matters because a ChatGPT user and a developer using an API model may face different product changes and deadlines.
The source also reports other scheduled platform changes. OpenAI’s Evals platform, Agent Builder and reusable prompt objects at v1/prompts are due to shut down November 30, 2026, with existing evals made read-only October 31. Later model retirements listed include tts-1 on January 6, 2027, older audio and realtime models on January 20, Whisper and specified transcription models on February 26, and gpt-5.1, gpt-5.4-nano and gpt-5.3-codex on April 1.
A separate change discussed in the source is the October 29 revision of OpenAI’s Model Spec, its public guide to model behavior. The source says the revision removes much of the detailed style guidance, including a default preference for Markdown, and asks models to interpret instructions by intent. That is not the same as an API shutdown, but it may matter to applications that rely on consistent formatting. The source also says OpenAI acknowledges its models “do not yet fully reflect” the Spec.
Notice Period and Migration Details
The source says OpenAI promises at least six months’ notice for generally available models except when safety or compliance concerns require faster action. It also says the GPT-4-family retirement was notified on October 14, three months before the January 14 date. The deprecation entry, as described in the source, does not say why that shorter interval applies. Developers with contractual commitments or reserved spending may want to ask their OpenAI account team about their specific arrangements.
OpenAI’s schedule reportedly says some customers may be able to buy dedicated capacity to retain access after a shutdown, but the source does not provide eligibility, terms or availability. It also does not establish whether the listed replacement models will match each retired model’s output on a particular application, or how individual workloads will change in cost. Prices and model availability can change, so developers should verify the current official pages before making decisions.
The source gives no independent benchmark results establishing equivalent quality or performance across the retired models and replacements. The Model Spec revision also does not by itself establish exactly how a particular application’s responses will change. Those effects remain workload-specific and unconfirmed until tested.
Audit and Test Before Cutoffs
Developers should start by searching application code, configuration files, automation tools and test fixtures for every affected model ID, including dated snapshots and fine-tuned variants. Calls using IDs retired October 23 need attention immediately; teams using GPT-5-family models face the December 11 deadline, followed by the specified GPT-4-family snapshots on January 14.
Before switching production traffic, run representative prompts through the listed replacement and compare correctness, formatting, latency and token use. Check whether sampling parameters are supported, whether output limits leave room for an answer, and how hidden reasoning billed as output affects costs. Applications that require JSON, Markdown or another fixed structure should state that format explicitly and validate the result.
The source says OpenAI documents deprecations and emails affected customers. Teams should check the live deprecations and pricing pages before each migration, confirm any account-specific options with OpenAI, and repeat tests as deadlines approach. The next listed date affecting model access is December 11, 2026; the January 14, 2027, GPT-4-family cutoff follows.
Key Questions
When does OpenAI retire GPT-5-nano from the API?
OpenAI’s schedule, as reported in the source, sets December 11, 2026 for GPT-5-nano. The listed replacement is gpt-5.6-luna.
When are the GPT-4o and GPT-4.1 API snapshots scheduled to stop working?
The specified GPT-4o, GPT-4o mini, GPT-4.1 and GPT-4.1 mini snapshots are scheduled for retirement on January 14, 2027. OpenAI lists gpt-6-luna as their replacement.
Have any of the affected models already been shut down?
Yes. The source says several IDs, including gpt-4.1-nano, gpt-4-turbo, gpt-3.5-turbo, o1 and o4-mini, were scheduled to shut down on October 23, 2026. It advises checking code for those IDs because calls may already fail.
Will migration necessarily reduce API costs?
No. The listed prices vary by model. The source reports lower short-context rates for gpt-6-luna than GPT-4o and GPT-4.1, but higher listed rates for gpt-5.6-luna than GPT-5-nano. Actual costs depend on token use, the applicable pricing and how the replacement handles a workload.
What should developers test before changing model IDs?
Test representative prompts for quality, formatting, latency and token costs. Review unsupported sampling parameters and output limits, and confirm that any required response format is explicitly requested. Verify current prices and retirement dates with OpenAI’s official pages.
Source: ThorstenMeyerAI.com
