News
ReleasesOVHcloud unveils Olala mini, its first in-house LLMChinaA 9B model beats a 1T model on FinQA, trained with Crusoe Managed RLReleasesThe internal metaphysics of an LLMReleasesMapping Language Model Internal Representations to Physical Laws and Scientific ReasoningResearchCan China keep rogue AI in check? Meet the start-up testing the top modelsFrontierMicrosoft aims to challenge OpenAI, Anthropic with cheaper, specialized AI modelsOpen modelsMicrosoft rolls out new first-party models and updates to agentic AI continuous improvementFrontierMicrosoft unveils AI models for coding, cybersecurity as it takes on Anthropic, OpenAIFrontierGoogle is making it easier to opt out of Gemini AI model training – here’s howFrontierOpenAI Unveils New Astra AI Model After Scrapping Prior ReleaseReleasesOVHcloud unveils Olala mini, its first in-house LLMChinaA 9B model beats a 1T model on FinQA, trained with Crusoe Managed RLReleasesThe internal metaphysics of an LLMReleasesMapping Language Model Internal Representations to Physical Laws and Scientific ReasoningResearchCan China keep rogue AI in check? Meet the start-up testing the top modelsFrontierMicrosoft aims to challenge OpenAI, Anthropic with cheaper, specialized AI modelsOpen modelsMicrosoft rolls out new first-party models and updates to agentic AI continuous improvementFrontierMicrosoft unveils AI models for coding, cybersecurity as it takes on Anthropic, OpenAIFrontierGoogle is making it easier to opt out of Gemini AI model training – here’s howFrontierOpenAI Unveils New Astra AI Model After Scrapping Prior Release
EveryLLM
imgOpenAI o4-mini

OpenAI o4-mini

Multimodal reasoning model by OpenAI, now retired.

OpenAI o4-mini was a generative pre-trained transformer designed for reasoning tasks. Released in April 2025, it processed text and images before being retired in February 2026.

Share on WhatsApp

  • Released to all ChatGPT users on 24 April 2025
  • Capable of analysing whiteboard sketches
  • Retired from ChatGPT on 13 February 2026
  • Included a slower, higher-effort o4-mini-high variant

What it is

OpenAI o4-mini was a generative pre-trained transformer focused on reasoning capabilities. Developed by OpenAI, it launched in April 2025 as a vision-language model. It could process both text and images, allowing it to analyse visual inputs like whiteboard sketches during its reasoning phase. The model served free and paid ChatGPT users, as well as developers via API, before its retirement in early 2026.

How it came about

Sam Altman originally intended to integrate this technology into GPT-5 rather than release it separately. However, OpenAI launched o4-mini alongside o3 on 16 April 2025 for paid subscribers and API access. A week later, on 24 April, it became available to all ChatGPT users, including those on the free tier. On 29 January 2026, OpenAI announced its retirement, which took effect on 13 February 2026.

Inside the model

The model functioned as a generative pre-trained transformer with multimodal capabilities. Unlike earlier iterations, it accepted image inputs alongside text. A key feature was its ability to perform chain-of-thought reasoning on visual data, such as interpreting whiteboard sketches. OpenAI also offered an o4-mini-high variant for paid users, which used more reasoning effort for greater accuracy at the cost of speed.

What it can do

o4-mini specialised in reasoning tasks and could process both text and images. It supported the analysis of visual elements like whiteboard drawings during its internal reasoning process. The high variant provided more accurate responses by dedicating additional computational effort to reasoning, though it operated more slowly than the standard version.

How to use it

The model was accessible via the Chat Completions API and Responses API. It was initially released to paid subscribers before expanding to all ChatGPT users, including free-tier accounts, on 24 April 2025. The o4-mini-high version remained exclusive to paid users. Access ended when the model was retired from ChatGPT on 13 February 2026.

Questions people ask about OpenAI o4-mini

What made o4-mini different from previous models?

Unlike earlier models, o4-mini could process both text and images. It specifically supported tasks like analysing whiteboard sketches during its chain-of-thought reasoning phase.

Is o4-mini still available today?

No. OpenAI announced its retirement on 29 January 2026, and it was fully removed from ChatGPT on 13 February 2026 alongside other legacy models.

What was the o4-mini-high model?

It was a variant available only to paid-tier users. This model was slower but applied higher reasoning effort to generate more accurate responses than the standard version.

Good reading and sources

Where this page comes from

The facts on this page come from the Wikipedia article "OpenAI o4-mini", written by its contributors and used under the CC BY-SA 4.0 licence, and from its Wikidata record (CC0). EveryLLM rewrote the text in its own words, with AI assistance, and compares every page with the article to make sure no passage is copied. It may be reused under the same CC BY-SA 4.0 licence with a link back.

It may contain mistakes, and figures change over time. Nothing here is investment or buying advice. If something is wrong, tell us. Last refreshed 2026-11-15.

The weekly digest

Every Monday: the model of the day, new pages on EveryLLM, and the conferences and anniversaries of the week ahead.