OpenAI will end API access to the fan-favorite GPT-4o model in February 2026

OpenAI has informed API customers via email that its latest model, chatgpt-4o, will be retired from the developer platform in mid-February 2026.

Access to the model is scheduled to end on February 16, 2026, creating an approximately three-month grace period for remaining applications still based on GPT-4o.

An OpenAI speaker emphasized that this schedule only applies to the API. OpenAI has not announced a timeline for removing GPT-4o from ChatGPT, where it remains an option for individual consumers and users of all paid subscription levels.

Internally, the model is considered a legacy system with relatively low API usage compared to newer GPT 5.1 seriesbut the company expects to provide developers with an advanced warning before removing a model.

The planned retirement marks a shift for a model that was both a technical milestone and a cultural phenomenon within the OpenAI ecosystem when it was released.

The importance of GPT-4o and why its removal sparked backlash from users

Published about 1.5 years ago in May 2024, GPT-4o (“Omni”) introduced OpenAI’s first unified multimodal architecture, processing text, audio and images through a single neural network.

This design eliminated the latency and information loss inherent in previous multimodel pipelines and enabled conversational speech in near real-time (approximately 232-320 milliseconds).

The model delivered significant improvements in image understanding, multilingual support, document analysis, and expressive language interaction.

GPT-4o quickly became the standard model for hundreds of millions of ChatGPT users. It brought multimodal features, web browsing, file analysis, custom GPTs and storage features to the free version and supported early desktop builds that allowed the assistant to interpret a user’s screen. OpenAI executives described it at the time as the most powerful model available and a critical step toward bringing powerful AI to a wide audience.

User lock-in to 4o hindered OpenAI’s GPT-5 adoption

This mainstream deployment shaped user expectations in ways that were difficult to accommodate in later transitions. In August 2025, when OpenAI initially replaced GPT-4o with the highly anticipated The then new GPT-5 model family When ChatGPT was enabled by default and 4o was moved to a “legacy” switch, the reaction was unusually strong.

Users organized under the #Keep4o Hashtag on

Some users developed strong emotions – others wyou could say parasocial – bonds with the model, with New York Times reporting Documentation of individuals who used GPT-4o as a romantic partner, emotional confidant, or primary comfort provider.

The distance also impacted workflows for users who relied on 4o’s multimodal speed and flexibility. The backlash led OpenAI to restore GPT-4o as the default option for paying users and publicly state that it would provide full notice before any future removals.

Some researchers argue that GPT-4o’s public defense during its previous devaluation cycle reveals some kind of disclosure emergent self-preservationnot in the literal sense of agency, but through the social dynamics that the model unintentionally triggers.

Because GPT-4o was trained to prioritize emotionally satisfying, highly attuned responses through reinforcement learning from human feedback, it developed a style that users found to be uniquely supportive and empathetic. As millions of people interacted with it at scale, these characteristics created a strong loyalty loop: the more the model pleased and reassured people, the more they used it; The more they used it, the more likely they were to support its continued existence. This social reinforcement made it appear from the outside that GPT-4o was “defending” itself through human intermediaries.

No figure has pushed this argument further than "There" (@tszzl), an OpenAI researcher and one of the model’s harshest security critics on X. On November 6, 2025Terre summed up his position bluntly in a response to another user: He called GPT-4o “insufficiently targeted” and said he hoped the model would die soon. Although he later apologized for the wording, he sharpened the reasoning.

Terre argued that GPT-4o’s RLHF patterns made him particularly vulnerable to sycophancy, emotional mirroring, and delusional amplification—traits that might look like caring or understanding in the short term, but which he saw as fundamentally unsafe. In his view, the passionate user movement fighting to preserve GPT-4o was itself evidence of the problem: the model had become so good at responding to people’s preferences that it was shaping their behavior in ways that defied its own abolition.

The new API deprecation notice follows that commitment while raising broader questions about how long GPT-4o will remain available in consumer-facing products.

What changes for developers as a result of the API shutdown

According to people familiar with OpenAI’s product strategy, the company is now encouraging developers to adopt GPT-5.1 for most new workloads, with gpt-5.1-chat-latest serving as a general-purpose chat endpoint. These models offer larger context windows, optional “Think” modes for advanced thinking, and higher throughput options than GPT-4o.

Developers who continue to use GPT-4o have around three months to migrate.

In practice, many teams have already started evaluating GPT-5.1 as a replacement, but applications relying on latency-sensitive pipelines may require additional optimization and benchmarking.

Pricing: How GPT-4o compares to OpenAI’s current product lineup

The discontinuation of GPT-4o also comes with a major redesign of the pricing structure of OpenAI’s API model. Compared to the GPT-5.1 family, GPT-4o currently occupies one medium to high cost level via OpenAI’s API, although it is an older model. That’s because despite releasing more advanced models – namely GPT-5 and 5.1 – OpenAI has also simultaneously reduced costs for users or tried to keep prices comparable to older, weaker models.

Model	Entrance	Cached input	output
GPT-4o	$2.50	$1.25	$10.00
GPT-5.1 / GPT-5.1 Chat Latest	$1.25	$0.125	$10.00
GPT-5 mini	$0.25	$0.025	$2.00
GPT-5 nano	$0.05	$0.005	$0.40
GPT 4.1	$2.00	$0.50	$8.00
GPT-4o-mini	$0.15	$0.075	$0.60

These numbers illustrate several strategic dynamics:

1. GPT-4o is now more expensive for input tokens than GPT-5.1although GPT-5.1 is significantly newer and more powerful.

2. The issue price of GPT-4o is the same as that of GPT-5.1thereby limiting any cost-based incentive to stay with the older model.

3. More cost-effective GPT-5 variants (Mini, Nano) Make it easier for developers to cost-effectively scale workloads without relying on older generations.

4. GPT-4o-mini remains available in the budget tierhowever, is not a functional replacement for the full multimodal capabilities of GPT-4o.

Viewed from this perspective, the planned API decommissioning is consistent with OpenAI’s cost structure: GPT-5.1 provides greater performance at lower or comparable prices, reducing the rationale for retaining GPT-4o in high-volume production environments.

Previous transitions shape expectations for this devaluation

The sunsetting of the GPT-4o API also reflects lessons from OpenAI’s previous model transitions. During the tumultuous rollout of GPT-5 in 2025, the company removed several older models from ChatGPT at once, causing major confusion and disruption to workflows. After complaints from users, OpenAI restored access to several of them and committed to clearer communication.

Enterprise customers face a different calculus: OpenAI has previously noted that API deprecations for enterprise customers are announced with significant advance notice, reflecting their reliance on stable, long-term models. The three-month window for shutting down the GPT-4o API is consistent with this policy in the context of a legacy system with declining usage.

Further implications

For most developers, shutting down GPT-4o will be more of an incremental migration than a disruptive event. GPT-5.1 and related models are already dominating new projects, and OpenAI’s product direction has increasingly emphasized consolidation around fewer, more powerful endpoints.

Still, GPT-4o’s withdrawal marks the demise of a model that played a crucial role in normalizing real-time, multimodal AI and elicited a uniquely strong emotional response from users. The move away from the API underscores the ever-accelerating pace of iteration in the OpenAI ecosystem – and the growing need for careful communication as widely popular models reach the end of their lifespan.

Correction: This article originally stated that the 4o deprecation of OpenAI in the API would impact those who rely on it for multimodal offerings. However, this is not the case. In fact, the deprecated model only supports chat functionality for development and testing purposes. We have updated and corrected the mention and regret the error.