🔍 Read the full analysis: What Developers Should Do About GPT-4o, GPT-4.1, And GPT-5-nano Leaving The API on ThorstenMeyerAI.com
Get business pricing on monitors, keyboards and dev gear
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
OpenAI’s API deprecation schedule lists several model shutdowns, including GPT-5, GPT-5 mini and GPT-5-nano on Dec. 11, 2026, and GPT-4o and GPT-4.1 snapshots on Jan. 14, 2027. Developers should inventory model IDs, verify replacement pricing and behavior, and test migrations; some older model IDs stopped working on Oct. 23.
OpenAI’s API deprecation schedule sets Dec. 11, 2026, for the shutdown of GPT-5, GPT-5 mini and GPT-5-nano, followed by GPT-4o and GPT-4.1 snapshots on Jan. 14, 2027. Developers using those model IDs will need to migrate their applications or risk failed API calls; several older IDs have already been retired, according to the schedule checked on Nov. 1.
The Dec. 11 list includes gpt-5, gpt-5-mini, gpt-5-nano, gpt-5-pro, o3 and o3-pro. OpenAI lists GPT-5.6 variants as replacements: GPT-5 to gpt-5.6-sol, GPT-5 mini to gpt-5.6-terra, and GPT-5-nano to gpt-5.6-luna. It also maps o3 to gpt-5.6-sol and o3-pro and GPT-5-pro to that model with its pro reasoning mode.
On Jan. 14, the scheduled retirements include dated snapshots of GPT-4o, GPT-4o mini, GPT-4.1 and GPT-4.1 mini. The recommended replacement listed for these is gpt-6-luna. The source says earlier IDs—including gpt-4.1-nano, the original GPT-4o snapshot, gpt-4-turbo, gpt-3.5-turbo, o1 and o4-mini—were shut down on Oct. 23. Calls still using those IDs may already fail.
Pricing and implementation details need checking before teams switch. The source’s Nov. 1 comparison, based on standard short-context prices per million tokens, lists GPT-4o at $2.50 input and $10 output, versus $0.10 and $0.50 for gpt-6-luna. By contrast, it lists GPT-5-nano at $0.05 input and $0.40 output, against $0.20 and $1.20 for its replacement, gpt-5.6-luna. Those prices are a dated snapshot, not a guarantee of future rates.
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 Production Apps
A model name in a configuration file, test fixture, or no-code automation can keep an older integration running unnoticed until the service stops accepting calls. Teams should search beyond application code: shared prompts, scheduled jobs, vendor tools and staging environments may also contain retired IDs. A failed request can interrupt a customer-facing feature or a background workflow.
Changing the model name alone may not preserve results. The source describes the replacements as reasoning models, for which hidden reasoning tokens are billed as output. It also warns that some sampling parameters need to be removed and that a tight output-token limit can lead to an empty, billed response. Those points make a migration a behavior and cost change, not just a configuration edit.
The listed prices also vary sharply by model pair. The GPT-4o-to-gpt-6-luna comparison is cheaper at the quoted rates, while the GPT-5-nano replacement costs more per token. Actual bills depend on traffic, prompt and output lengths, caching, and the price schedule in effect when the migration happens. Teams should model their own usage rather than assume every replacement reduces costs.
Products Worth Considering
Logitech MX Mechanical Wireless Illuminated Keyboard, Tactile Quiet, Graphite
Logitech MX Keys S Wireless Keyboard, Graphite
Samsung 49-Inch Odyssey G93SC QD-OLED Curved Gaming Monitor
As an affiliate, we earn on qualifying purchases.
Dates Behind the API Changes
The source says OpenAI’s deprecations page records affected models and that the company emails customers. It groups the schedule into several waves: some models stopped working on Oct. 23, 2026; GPT-5-family models and o3 are listed for Dec. 11; and the named GPT-4o and GPT-4.1 snapshots are listed for Jan. 14, 2027. Other audio, transcription and coding models have dates later in 2027.
The notice distinguishes dated snapshot IDs from shorter model names commonly used in code. The source reports that the model pages for the short names gpt-4o, gpt-4o-mini, gpt-4.1 and gpt-4.1-mini show a deprecated status and says developers should treat them as covered by the January date. Teams should verify the current entry for every ID they use rather than rely on a name that appears in a guide or an old configuration.
The API schedule is separate from a ChatGPT product change. According to the source, OpenAI removed GPT-4o, GPT-4.1 and GPT-4.1 mini from ChatGPT on Feb. 13, 2026; the current notice concerns API access and applications built on it. The source also reports that OpenAI’s Evals platform, Agent Builder and reusable prompt objects are scheduled to shut down on Nov. 30, with existing evals read-only from Oct. 31.
““If your app depends on a format, say so in the request.””
— Thorsten Meyer, ThorstenMeyerAI.com
Pricing and Notice Details to Verify
The schedule and price comparisons here come from a source checked on Nov. 1, 2026; prices and availability can change. The source says OpenAI’s deprecation page promises at least six months’ notice for generally available models unless safety or compliance concerns require a faster timeline. It also notes that the Jan. 14 retirements were listed on Oct. 14, a three-month interval, but says the notice does not explain which exception or rule applies. The reason for that timing is not established in the supplied material.
It is also unclear whether any particular customer will receive continued access after the listed shutdowns. The source says OpenAI may offer some customers dedicated capacity, but gives no eligibility criteria or terms. Developers should confirm their own account arrangements and the exact shutdown status of each model ID with OpenAI.
The source summarizes behavior and pricing risks but does not provide independent migration test results. It does not establish that every application will produce different answers, fail, or cost more after switching. Those outcomes depend on the model, request settings, workload and updated prices.
Audit, Test, Then Switch
Start with an inventory of model IDs across source code, environment variables, prompt templates, test fixtures, automation platforms and third-party services. Flag IDs listed as already shut down, then prioritize systems using models scheduled for Dec. 11 and Jan. 14. Assign an owner and deadline to each affected integration, and check OpenAI’s current deprecations page and account notices rather than treating this schedule as static.
For each replacement, run representative production-like requests before changing live traffic. Compare output quality and formatting, latency, error rates, token use and total cost. Test edge cases, including long prompts, structured responses and requests with tight output limits. Review request parameters against the replacement’s requirements, and make format instructions explicit if the application relies on a particular structure.
Where possible, roll out the new model in a controlled way, monitor results, and keep a documented rollback plan that does not depend on a model after its retirement date. Recheck prices and shutdown dates before each migration; the next listed milestones are the Nov. 30 tooling shutdown, Dec. 11 model retirements and Jan. 14, 2027, GPT-4o and GPT-4.1 snapshot retirements.
Key Questions
When are GPT-5-nano and the other GPT-5 models scheduled to leave the API?
OpenAI’s schedule, as summarized in the source checked Nov. 1, 2026, lists Dec. 11, 2026, for gpt-5, gpt-5-mini, gpt-5-nano, gpt-5-pro, o3 and o3-pro.
When are GPT-4o and GPT-4.1 API snapshots scheduled to stop working?
The listed date is Jan. 14, 2027, for specified GPT-4o, GPT-4o mini, GPT-4.1 and GPT-4.1 mini snapshots. The source advises checking the deprecation status of short model names as well as dated IDs.
What should developers test before switching models?
Test representative prompts for output quality, formatting, latency, errors, token use and cost. Review request parameters and output limits because the source says the replacements use reasoning and may require parameter changes; a tight output limit can produce an empty, billed answer.
Will every replacement cost less?
No. The source’s Nov. 1 price comparison lists lower prices for some pairings, including GPT-4o to gpt-6-luna, and higher per-token prices for GPT-5-nano to gpt-5.6-luna. Costs depend on actual usage and the prices in effect at migration time.
Are these retirements the same as models leaving ChatGPT?
No. The dates in this report concern the API and applications built with it. The source separately reports that OpenAI removed GPT-4o, GPT-4.1 and GPT-4.1 mini from ChatGPT on Feb. 13, 2026.
Source: ThorstenMeyerAI.com
