🔍 Read the full analysis: GPT-4o And GPT-4.1 API Users: What The OpenAI Switch-Off Means 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 deprecation schedule says GPT-4o, GPT-4o mini, GPT-4.1 and GPT-4.1 mini API snapshots are due to stop working on Jan. 14, 2027; several GPT-5 and older models have earlier deadlines. The listed replacements change pricing and model behavior, so API users should check model IDs, test outputs and verify current prices before migrating.
OpenAI plans to retire the listed GPT-4o and GPT-4.1 API snapshots on January 14, 2027, requiring developers using those model IDs to move to a replacement or risk failed requests after the deadline. The company’s schedule also lists several GPT-5 models for retirement on December 11, 2026, while a separate group of older API models was scheduled to stop working on October 23, 2026.
According to OpenAI’s deprecations page, the January deadline 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. The listed recommended replacement is GPT-6 Luna. The source report says the short model names are also marked deprecated, so developers should check both names in application settings and dated IDs in code.
The earlier December 11, 2026, schedule covers GPT-5, GPT-5 mini, GPT-5 nano, GPT-5 pro, o3 and o3-pro, with replacement models listed for each. The report says some older models—including GPT-4 Turbo, GPT-3.5 Turbo, GPT-4.1 nano, the original GPT-4o snapshot, o1 and o4-mini—were already scheduled to shut down on October 23. Applications still calling those IDs may already be affected; the report does not establish how many users or requests are impacted.
OpenAI’s listed replacements do not all have the same per-token prices as the retiring models. The source’s comparison, based on standard short-context rates checked on November 1, says GPT-6 Luna is listed at $0.10 per million input tokens and $0.50 per million output tokens, compared with GPT-4o at $2.50 and $10.00. By contrast, GPT-5 nano’s listed replacement, GPT-5.6 Luna, is priced at $0.20 input and $1.20 output, compared with $0.05 and $0.40 for GPT-5 nano. These are stated rates, not estimates of a particular customer’s total bill.
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 deadlines matter because an API model ID can be embedded far beyond a main application’s visible settings: in configuration files, tests, scheduled jobs and third-party automation. If a request continues to use a retired ID after its shutdown, the application may fail rather than automatically adopting the replacement. Teams need to identify every use of a model name and plan a change before the applicable date.
A replacement can also change costs and outputs even when an application continues to run. The source report characterizes the listed replacements as reasoning models and warns that hidden reasoning tokens are billed as output. It also says some sampling parameters must be removed and a restrictive output-token limit can result in an empty answer that is still billed. Those operational details are claims in the report; teams should verify them against current OpenAI model documentation and test their own workloads.
For product owners, this is not only a model-name update. Changes to response format, latency, token use or behavior can affect user-facing features and automated workflows. A migration should be evaluated against the application’s required output and budget, not just whether a replacement accepts the same request.
Products Worth Considering
LG UltraGear 32GX850A-B 32-Inch 4K UHD Glossy OLED Gaming Monitor
LG UltraGear 27-Inch QHD OLED Gaming Monitor, 240Hz, 0.03ms, G-SYNC Compatible, FreeSync Premium Pro, HDMI 2.1
Samsung 27 Odyssey OLED G5 (G50SF) QHD QD-OLED Gaming Monitor
As an affiliate, we earn on qualifying purchases.
OpenAI’s Published Retirement Timeline
The dates and model lists in the source report are drawn from OpenAI’s deprecations page, which its author says was checked on November 1, 2026. The report says OpenAI notifies affected customers by email and documents deprecations on that page. Its schedule also includes other products and models: the Evals platform, Agent Builder and reusable prompt objects are listed for November 30, 2026, with existing evals described as read-only from October 31.
The API schedule is separate from changes to ChatGPT. The report says GPT-4o, GPT-4.1 and GPT-4.1 mini were removed from ChatGPT on February 13, 2026. That earlier consumer-product change does not itself mean API access ended at the same time; the current deadlines concern API model IDs used by developers and services.
The source also reports that OpenAI’s Model Spec was revised on October 29, 2026, with much of its detailed style guidance removed, including a default for Markdown formatting. That policy document is distinct from the API retirement notice, but it may matter to teams whose applications depend on a particular response style. Developers can state required formats in their prompts and test whether the replacement follows them.
Pricing and Migration Questions
The schedule sets out planned dates and recommended replacements, but the source report does not say whether OpenAI will change those dates or whether every customer has received a notice. It also says the stated general policy provides at least six months’ notice for generally available models, subject to safety or compliance exceptions, while the GPT-4o and GPT-4.1 notice-to-retirement period described here is three months. The reason for that timing is not specified in the report.
Prices can change, and the figures cited are standard short-context rates from a November 1 check, not a customer-specific quote. Actual costs depend on usage and billing details. The report says some customers may be able to buy dedicated capacity for continued access after a shutdown, but it does not provide eligibility rules, availability or pricing. Fine-tuned versions are said to follow their base models’ retirement schedules; the replacement models listed in the report do not support fine-tuning. Developers with fine-tuned deployments should confirm their options directly with OpenAI.
It is also not clear from the source material how the replacements will perform on individual applications, or whether behavior changes attributed to the revised Model Spec will affect a given API workflow. Those questions require current product documentation and application-level testing.
Audit Models Before Each Deadline
API users should first inventory model IDs across source code, configuration, test fixtures, automation tools and deployed services. Check for both short names and dated snapshots, then compare each one with OpenAI’s current deprecation page. Any service using a model scheduled for October 23 should be investigated immediately; GPT-5 users should work to a December 11 deadline, while listed GPT-4o and GPT-4.1 snapshot users should plan around January 14.
Next, test the recommended replacement with representative real-world requests. Compare answer quality, formatting, latency, token consumption, error handling and total cost. Review parameters that may not be supported and output limits that could affect reasoning-model responses. Run the tests in a staging environment, update affected integrations and monitor production after deployment. OpenAI’s schedule and prices should be rechecked before the switch, since both may be updated.
Organizations with committed spend, fine-tuned models or a need for continued access should contact their OpenAI account team to confirm options, notice terms and any dedicated-capacity availability. The next published milestones in the source report are the November 30 shutdown of several development tools, the December 11 model retirements and the January 14 GPT-4o and GPT-4.1 API retirements.
Key Questions
When are GPT-4o and GPT-4.1 API models scheduled to stop working?
OpenAI’s schedule, as reported in the source, lists January 14, 2027, for the specified GPT-4o, GPT-4o mini, GPT-4.1 and GPT-4.1 mini snapshots. Check the current deprecations page for any changes.
Are GPT-5 API models being retired on the same date?
No. The source report lists December 11, 2026, for GPT-5, GPT-5 mini, GPT-5 nano, GPT-5 pro, o3 and o3-pro. It gives January 14, 2027, for the listed GPT-4o and GPT-4.1 snapshots.
What replacement does OpenAI list for the GPT-4o and GPT-4.1 snapshots?
The source report says OpenAI lists GPT-6 Luna as the recommended replacement for the covered GPT-4o and GPT-4.1 snapshots. Developers should confirm current guidance and test compatibility before changing production applications.
Will the replacement cost less?
Not for every workload. The report’s November 1 short-context price comparison lists GPT-6 Luna below GPT-4o and GPT-4.1 on standard input and output rates, but lists some GPT-5 replacements at higher per-token rates than their predecessors. Total cost depends on actual usage, and prices may change.
What should developers do before the shutdown?
Find all uses of affected model IDs, check OpenAI’s latest schedule, test the replacement with representative requests, review pricing and supported parameters, and deploy changes before the deadline. Teams using fine-tuned models or requiring continued access should ask OpenAI about their specific options.
Source: ThorstenMeyerAI.com
