OpenAI

Maker of GPT and ChatGPT

https://openai.com

News

analysisOpenAI

OpenAI reasoning model solves 80-year math problem as Anthropic hits $10.9B quarterly revenue

In a two-hour span Wednesday, OpenAI announced its reasoning model autonomously solved an 80-year-old geometry problem while Anthropic reported it's on track for $10.9 billion in Q2 revenue with $559 million in operating profit—two years ahead of internal projections. The developments came alongside Nvidia's $81.6 billion quarter, Anthropic's $1.25 billion monthly SpaceX compute deal, and a White House AI executive order signing.

2 min read
researchOpenAI

OpenAI claims reasoning model disproved 80-year-old Erdős conjecture in geometry

OpenAI claims its new reasoning model has produced an original mathematical proof disproving a geometry conjecture first posed by Paul Erdős in 1946. The company says this is the first time AI has autonomously solved a prominent open problem central to a field of mathematics, with verification from mathematicians including Thomas Bloom and Noga Alon.

2 min read
product updateOpenAI

OpenAI builds custom Windows sandbox for Codex coding agent after existing tools proved insufficient

OpenAI has implemented a custom sandbox for its Codex coding agent on Windows after determining that existing Windows isolation tools—AppContainer, Windows Sandbox, and Mandatory Integrity Control—could not adequately balance safety and functionality. The solution uses synthetic SIDs and write-restricted tokens to constrain file writes and network access without requiring administrator privileges.

2 min read
product updateOpenAI

OpenAI launches Daybreak cybersecurity platform with GPT-5.5 variants for vulnerability detection

OpenAI has launched Daybreak, a cybersecurity platform built on three GPT-5.5 model variants designed to detect software vulnerabilities, generate patches, and validate fixes in enterprise codebases. The platform directly competes with Anthropic's Mythos and includes partnerships with eight major security companies including Cisco, Cloudflare, and CrowdStrike.

2 min read
product updateOpenAI

OpenAI launches Trusted Contact feature allowing ChatGPT to alert designated friends during suicide risk

OpenAI has launched Trusted Contact for ChatGPT, allowing users 18+ to designate one adult contact who can be notified if the company's trained human review team detects serious self-harm risk. The feature comes after over 1 million of ChatGPT's 800 million weekly users expressed suicidal thoughts in conversations, and follows a 2025 wrongful death lawsuit.

2 min read
product updateOpenAI

OpenAI launches GPT-Realtime-2 with GPT-5-class reasoning, adds real-time translation across 70 languages

OpenAI has added three voice intelligence features to its Realtime API: GPT-Realtime-2 with GPT-5-class reasoning for complex conversational requests, GPT-Realtime-Translate supporting 70 input languages and 13 output languages, and GPT-Realtime-Whisper for live speech-to-text transcription. Translation and transcription are billed by the minute, while GPT-Realtime-2 uses token-based pricing.

2 min read
model releaseOpenAI

OpenAI Opens GPT-5.5-Cyber to Vetted Defenders After Model Matches Anthropic's Mythos in Security Testing

OpenAI is providing a less-restricted version of GPT-5.5 to vetted cybersecurity defenders through its Trusted Access for Cyber program. The model, dubbed GPT-5.5-Cyber, completed a 32-step simulated corporate cyberattack in 2 out of 10 test runs according to the U.K. AI Security Institute, narrowly trailing Anthropic's Mythos which succeeded in 3 out of 10 attempts.

2 min read

Models

GPT-2 774M

OpenAI

discontinued
Context1K

GPT-3

OpenAI

discontinued
Context4K

OpenAI Codex (2024)

OpenAI

discontinued

OpenAI Codex (codex-1)

OpenAI

discontinued

Point-E

OpenAI

discontinued

GPT-4

OpenAI

discontinued

Voice Engine

OpenAI

discontinued

GPT-3.5 Turbo

OpenAI

discontinued
Context4K
Input/1M$2

GPT-Realtime-2

OpenAI

active
Input/1M$32

May 7, 2026

GPT-5.5-Cyber

OpenAI

active

May 7, 2026

GPT-5.5-Cyber

OpenAI

active

May 7, 2026

GPT-5.5 Instant

OpenAI

active

May 5, 2026

GPT-5.5

OpenAI

active
Context1000K
Input/1M$5

Apr 29, 2026

OpenAI GPT Mini Latest

OpenAI

active
Context400K

Apr 27, 2026

GPT-5.5 Pro

OpenAI

active
Context1050K
Input/1M$30

Apr 24, 2026

OpenAI Privacy Filter

OpenAI

active
Context128K

Apr 22, 2026

ChatGPT Images 2

OpenAI

active

Apr 21, 2026

ChatGPT Images 2.0

OpenAI

active

Apr 21, 2026

ChatGPT Images 2.0

OpenAI

active

Apr 21, 2026

GPT-5.4 Image 2

OpenAI

active
Context272K
Input/1M$8

Apr 21, 2026

GPT-5.4-Cyber

OpenAI

active

Apr 15, 2026

GPT-5.4 Nano

OpenAI

active

The most lightweight and cost-efficient variant of the GPT-5.4 family. Optimized for speed-critical and high-volume tasks. Supports text and image inputs for classification, data extraction, and agent execution.

Context400K
Input/1M$0.2

Mar 17, 2026

GPT-5.4 mini

OpenAI

active
Context400K
Input/1M$0.75

Mar 17, 2026

GPT-5.4

OpenAI

active
Context1049K
Input/1M$2.5

Mar 5, 2026

DALL·E

OpenAI

discontinued

Feb 19, 2026

DALL·E 3

OpenAI

discontinued

Feb 19, 2026

o4-mini

OpenAI

active

Fast, affordable reasoning model. Strong on STEM tasks.

Context200K
Input/1M$1.1

Apr 16, 2025

o3

OpenAI

active

OpenAI's most capable reasoning model. Leads on math, code, and science.

Context200K
Input/1M$10

Apr 16, 2025

GPT-4.1 mini

OpenAI

active

Efficient GPT-4.1 with 1M context. Best value for agentic and long-context tasks.

Context1000K
Input/1M$0.4

Apr 14, 2025

GPT-4.1

OpenAI

active

GPT-4.1 with 1M token context. Built for long-document and agentic workflows.

Context1000K
Input/1M$2

Apr 14, 2025

GPT-4.1 nano

OpenAI

active

Smallest and cheapest GPT-4.1. Optimized for classification and low-latency tasks.

Context1000K
Input/1M$0.1

Apr 14, 2025

GPT-Rosalind

OpenAI

active

Apr 10, 2025

GPT-4.5

OpenAI

active

OpenAI's largest GPT model. Strongest emotional intelligence and world knowledge of any GPT, optimized for agentic workflows.

Context128K
Input/1M$75

Feb 27, 2025

o3-mini

OpenAI

deprecated

Fast reasoning model at o3-level performance. Configurable thinking effort (low/medium/high).

Context200K
Input/1M$1.1

Jan 31, 2025

o1-pro

OpenAI

active

o1 with extended compute for the hardest problems. Significantly outperforms o1 on math, science, and complex coding tasks.

Context200K
Input/1M$150

Dec 17, 2024

o1

OpenAI

deprecated

OpenAI's flagship reasoning model with extended chain-of-thought. Strong on math and science.

Context200K
Input/1M$15

Dec 5, 2024

o1-mini

OpenAI

deprecated

Compact reasoning model. Superseded by o3-mini and o4-mini.

Context128K
Input/1M$3

Sep 12, 2024

GPT-4o August 6, 2024

OpenAI

active
Context128K
Input/1M$2.5

Aug 6, 2024

GPT-4o mini

OpenAI

active

Small, affordable GPT-4o for lightweight tasks.

Context128K
Input/1M$0.15

Jul 18, 2024

GPT-4o mini

OpenAI

active
Context128K
Input/1M$0.15

Jul 18, 2024

GPT-4o

OpenAI

active
Context128K
Input/1M$2.5

May 13, 2024

GPT-4 Turbo

OpenAI

deprecated

Original GPT-4 Turbo with 128K context. Superseded by GPT-4o.

Context128K
Input/1M$10

Apr 9, 2024

Top Benchmark Scores

Full leaderboard →

AIME 2024

GPT-5.4
94%

AIME 2025

o3
96.7%

DocVQA

GPT-5.4
97%
93.5%

HumanEval

o4-mini
99.4%

LiveCodeBench

GPT-5.4
83.5%
99.5%
93.8%

MMLU-Pro

GPT-5.4
89.8%
85.5%

Speed (tok/s)

GPT-5.4 Nano
210 tokens_per_sec

SWE-bench Verified

GPT-5.4
83.1%