🔍 Read the full analysis: Three OpenAI API Models Are Switching Off: What Developers Should Check on ThorstenMeyerAI.com
Prime made for students and young adults
- Fast, free delivery for dorm and study essentials
- Prime Video and Amazon Music included
- Member-only deals
TL;DR
OpenAI’s deprecation schedule lists API shutdowns on December 11, 2026, and January 14, 2027, with other models already retired on October 23. Developers should find affected model IDs, verify replacement pricing and behavior, and test changes before calls stop working. The source says replacement models may handle formatting and reasoning differently, but some migration details remain unclear.
OpenAI plans to retire GPT-5, GPT-5 mini, GPT-5 nano, o3 and related API models on December 11, 2026, followed by listed GPT-4o and GPT-4.1 snapshots on January 14, 2027, according to the company’s deprecation schedule as reported by ThorstenMeyerAI.com. The changes matter to developers because applications still calling affected model IDs may fail after the relevant date, while replacements can differ in price and response behavior.
The December 11 retirement list includes gpt-5, gpt-5-mini, gpt-5-nano, gpt-5-pro, o3 and o3-pro. The source says OpenAI recommends 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. The January 14 list covers dated snapshots of GPT-4o, GPT-4o mini, GPT-4.1 and GPT-4.1 mini; the source identifies gpt-6-luna as their recommended replacement.
Some models were already listed as shut down on October 23, 2026, including gpt-3.5-turbo, gpt-4-turbo, gpt-4.1-nano, the original GPT-4o snapshot, o1, o3-mini and o4-mini. Developers should check both explicit dated snapshot IDs and short model names in application code, configuration, test fixtures and automation tools. The source says affected fine-tuned models follow the retirement date of their base model.
Pricing comparisons in the source use standard short-context rates per million tokens, checked November 1. It lists GPT-4o at $2.50 per million input tokens and $10 per million output tokens, compared with gpt-6-luna at $0.10 and $0.50. The change is different for smaller models: the listed GPT-5 nano replacement costs $0.20 per million input tokens versus $0.05 for GPT-5 nano, while GPT-5 mini’s listed replacement costs $2 versus $0.25 per million input tokens. Prices and model availability should be checked against OpenAI’s current pages before migration decisions.
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 Users
A model retirement can affect more than whether an API request succeeds. A replacement may change cost, latency, output format or task performance, so switching the model name without testing could disrupt a production workflow. The source describes the recommended replacements as reasoning models, which may use hidden reasoning tokens billed as output and may require developers to revisit sampling settings and output limits.
The price changes are not uniform. The source’s figures show lower listed rates for some GPT-4 family migrations, but higher rates for the GPT-5 nano and GPT-5 mini replacements. Teams should estimate spend using their own input and output volumes rather than assuming a migration will reduce costs. The source also warns that an overly tight output limit can result in an empty answer while tokens are still billed; developers should verify this behavior in their own tests.
Behavior changes also matter for products that depend on predictable formatting. The source reports that a revised Model Spec dated October 29 removed much of its detailed style guidance, including a default preference for Markdown. That document is guidance about model behavior, not a guarantee that every application will change; teams should test their prompts and explicitly request formats their software needs.
Popular right now
Vitamix Propel Series 750 Professional-Grade Blender, 4 Programs, Black
Ninja CREAMi Scoop & Swirl 13-in-1 Ice Cream and Soft Serve Maker (NC701)
KitchenAid Artisan 5-Quart Tilt-Head Stand Mixer, Almond Cream
As an affiliate, we earn on qualifying purchases.
The Schedule Behind the Retirements
The source says its dates come from OpenAI’s deprecations page, checked on November 1, 2026, and that OpenAI emails affected customers and documents deprecations there. It reports that the December retirements were notified on June 11 and the January retirements on October 14. The source also says OpenAI’s general notice policy promises at least six months for generally available models, subject to faster action for safety or compliance reasons. It is not clear from the cited entry which notice rule applies to the January retirements.
Developers should distinguish the API schedule from changes to ChatGPT. According to the source, OpenAI removed GPT-4o, GPT-4.1 and GPT-4.1 mini from ChatGPT on February 13, 2026; the current dates concern API access used by external applications and services. The source also lists a separate November 30 shutdown for OpenAI’s Evals platform, Agent Builder and reusable prompt objects, with existing evaluations becoming read-only on October 31. Those dates may affect teams preparing migration tests.
Open Questions Before Migration
The supplied source does not include direct statements from OpenAI about why each retirement is scheduled, nor does it reproduce the full deprecation entries. Developers should confirm the affected IDs, replacement recommendations and dates on OpenAI’s live documentation before acting. The reported three-month interval between the October 14 notice and the January 14 shutdown also differs from the general six-month notice period described in the source; the reason for that difference is not established in the material provided.
It is also unclear whether every application will see a material change in outputs after migration. That depends on prompts, workloads, model settings and the replacement selected. The source says the listed replacements do not support fine-tuning, but provides no further details on migration options for customers whose fine-tuned models are affected. OpenAI’s dedicated-capacity arrangements, which the source says may be available to some customers, are not described in enough detail to establish eligibility or terms.
Checks to Run Before Deadlines
Developers should start by inventorying every model ID used across code, environment settings, scheduled jobs, third-party integrations and no-code workflows. Compare that list with OpenAI’s current deprecation page, including dated snapshots and short names, and identify the service owner for each call. For any model already retired on October 23, check logs for failed requests and replace the ID where needed.
For models scheduled to retire, test the recommended replacement against representative production prompts before the shutdown date. Compare output quality and required formatting, measure token use and latency, confirm whether existing parameters are accepted, and test output limits with realistic workloads. Recalculate expected costs using current rates and each model’s input, output and cached-token usage. Teams relying on fine-tuned versions or committed spend should ask OpenAI about their options directly. The next scheduled dates in the supplied material are November 30 for the listed platform tools, December 11 for GPT-5 and o-series retirements, and January 14 for the GPT-4 family snapshots.
Key Questions
Which API models are scheduled to retire on December 11, 2026?
The source lists gpt-5, gpt-5-mini, gpt-5-nano, gpt-5-pro, o3 and o3-pro. Check OpenAI’s current deprecation page to confirm the affected IDs and recommended replacements.
When are the listed GPT-4o and GPT-4.1 snapshots scheduled to stop working?
The reported date is January 14, 2027, for listed GPT-4o, GPT-4o mini, GPT-4.1 and GPT-4.1 mini snapshots. The source says gpt-6-luna is the recommended replacement for these models.
What should developers check first?
Search all code and configuration for affected model IDs, including dated snapshots and short names. Then verify current retirement dates, test the replacement’s output and settings, and estimate costs using your application’s actual token usage.
Will migration necessarily make API usage cheaper?
No. The source’s listed rates are lower for some GPT-4 family replacements but higher for the cited GPT-5 nano and GPT-5 mini replacements. Costs depend on the selected model and input and output token volumes.
What happens to fine-tuned versions of retiring models?
The source says fine-tuned models follow the retirement schedule of their base model and that the listed replacements do not support fine-tuning. It does not provide enough detail to determine what migration options are available for each affected customer.
Source: ThorstenMeyerAI.com
