🔍 Read the full analysis: GPT-4o, GPT-4.1, And GPT-5-nano API Changes: A Guide To Timing And Impact 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 published deprecation schedule says several API models have already shut down, with further retirements set for December 11, 2026, and January 14, 2027. The listed replacements can change costs and behavior, so developers should check model IDs, test applications and confirm the current schedule with OpenAI.
OpenAI’s API deprecation schedule sets December 11, 2026, for the retirement of GPT-5, GPT-5 mini and GPT-5-nano, followed by listed GPT-4o and GPT-4.1 model snapshots on January 14, 2027. Developers using those models will need to move to another model or risk failed API calls; the replacements listed by OpenAI may also change costs and application behavior.
According to OpenAI’s deprecations page, checked by ThorstenMeyerAI.com on November 1, a number of older model IDs have already been shut down. The October 23 list includes gpt-4.1-nano, the original GPT-4o snapshot, GPT-4 Turbo, GPT-3.5 Turbo, o1 and o4-mini, among others. Any application still calling a retired ID may fail. OpenAI says it notifies affected customers by email and lists deprecations on its website.
The next major date is December 11, 2026. The listed retirements include GPT-5, GPT-5 mini, GPT-5-nano, GPT-5-pro, o3 and o3-pro. OpenAI’s listed replacement mapping sends GPT-5-nano users to GPT-5.6 Luna and GPT-5 mini users to GPT-5.6 Terra; GPT-5 and o3 are mapped to GPT-5.6 Sol. The schedule then lists GPT-4o, GPT-4o mini, GPT-4.1 and GPT-4.1 mini snapshots for retirement on January 14, with GPT-6 Luna named as their replacement.
The replacement models are not necessarily cost-neutral or behaviorally identical. In the source’s comparison of standard prices per million tokens for prompts up to 272,000 tokens, GPT-4o is listed at $2.50 per million input tokens, compared with $0.10 for GPT-6 Luna. By contrast, GPT-5-nano’s listed replacement costs $0.20 per million input tokens versus $0.05 for the outgoing model. The source also reports that the replacements are reasoning models, for which reasoning tokens count as output and some sampling parameters may not be supported. Prices and model availability should be rechecked directly before migration.
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
The dates matter to teams whose software, internal tools or customer-facing services call a specific model ID. A retirement can turn a quiet configuration dependency into failed requests, while changing the model may affect latency, output format, quality and the bill. The impact depends on how each application uses the model; the announced replacement is not proof that a workload will behave the same.
Price comparisons also cut in different directions. The source’s listed rates show a large reduction in input price when moving from GPT-4o to GPT-6 Luna, but an increase for users moving from GPT-5-nano to GPT-5.6 Luna. Reasoning-token billing and output limits may further affect total spend. Teams should compare full task costs using their own traffic and test results, rather than relying on a single per-token figure.
There is a second operational issue: ThorstenMeyerAI.com says OpenAI’s October 29 Model Spec revision removes much of its prior style guidance, including a default preference for Markdown. The article reports that OpenAI advises models to interpret instructions by intent and says the models do not yet fully reflect the Spec. That document change does not itself confirm that a particular API response will change, but applications relying on implicit formatting conventions should specify required formats directly and validate the outputs.
Products Worth Considering
LG UltraGear 32GX850A-B 32-Inch 4K UHD Glossy OLED Gaming Monitor
LG 34GX90SA-W 34-inch Ultragear WQHD OLED Curved Gaming Monitor
Samsung 27-inch Odyssey OLED G6 QD-OLED QHD Gaming Monitor
As an affiliate, we earn on qualifying purchases.
Two API Changes on Separate Clocks
This is an API retirement update, not a notice that the same models are leaving every OpenAI product at once. The source says GPT-4o, GPT-4.1 and GPT-4.1 mini were removed from ChatGPT on February 13, 2026, while the dates at issue here concern access through the API, which developers use to build their own products and workflows.
The schedule extends beyond the two main deadlines. The source lists a November 30, 2026, shutdown for OpenAI’s Evals platform, Agent Builder and reusable prompt objects, with existing evals described as read-only from October 31. It also lists later 2027 retirements for some speech, audio, transcription and other models. The source says OpenAI’s page promises at least six months’ notice for generally available models unless safety or compliance concerns call for a faster timeline; it notes that the October 14 notice for the January 14 retirements is three months, and says the reason for that timing is not stated.
OpenAI’s deprecation table names dated snapshots, while developers may use shorter model names in their code. ThorstenMeyerAI.com advises treating the short GPT-4o and GPT-4.1 names as affected by the January date because their model pages are marked deprecated and their snapshots appear on the schedule. The article also says fine-tuned models follow their base model’s retirement and that the named replacements do not support fine-tuning.
“The replacements think before they answer.”
— Thorsten Meyer, ThorstenMeyerAI.com
Policy Timing and Workload Effects
The schedule and pricing details here are based on the source’s check of OpenAI pages on November 1, 2026; both may change. The source does not establish why the January retirements received three months’ notice despite the stated six-month policy for generally available models. It says OpenAI allows for shorter notice in some circumstances, but the schedule entry does not specify whether that exception applies.
It is also not clear from the source how each customer’s applications will perform after migration, what their final costs will be, or whether every account has the same replacement access. The source says some customers may be able to buy dedicated capacity to retain access after a shutdown, but provides no eligibility criteria or terms. Developers should confirm these points with OpenAI rather than assume a replacement is available on identical terms.
Audit Model IDs Before Deadlines
Teams should start by searching application code, configuration files, test fixtures and automation tools for every model ID, including dated snapshots and fine-tuned variants. Record which services depend on each ID, then check the current OpenAI deprecations and pricing pages and contact an account representative about any contractual or dedicated-capacity questions.
Before switching production traffic, test the proposed replacement against representative prompts and expected outputs. Check formatting, tool use, latency, quality, token consumption and total cost, including reasoning tokens where applicable. Review output-token limits and remove parameters that the new model does not support. Add explicit instructions for required formats instead of relying on an assumed default. The next scheduled API milestones are November 30 for the listed tools and December 11 for the GPT-5 and o3 models, followed by the January 14 GPT-4o and GPT-4.1 retirements.
Key Questions
When are GPT-5-nano and GPT-4o scheduled to leave the API?
According to the schedule cited by ThorstenMeyerAI.com, GPT-5-nano is listed for December 11, 2026. The listed GPT-4o snapshots are scheduled for January 14, 2027. Check OpenAI’s current deprecations page for any changes.
Does this mean the models are being removed from ChatGPT?
No. This schedule concerns the API. The source separately reports that GPT-4o, GPT-4.1 and GPT-4.1 mini were removed from ChatGPT on February 13, 2026.
Will switching models lower my API bill?
Not necessarily. The listed prices vary by model and token type: the source reports lower input pricing for GPT-4o users moving to GPT-6 Luna, but higher input and output rates for GPT-5-nano users moving to GPT-5.6 Luna. Actual costs depend on usage and should be tested against current prices.
What should developers test before migrating?
Test representative requests for output quality, formatting, latency, token use and cost. Check whether the replacement supports the parameters your application sends, review output limits and make required response formats explicit.
Source: ThorstenMeyerAI.com
