OpenAI Blog

OpenAI Welcomes Sarah Friar (CFO) and Kevin Weil (CPO)

Clip: Connecting Text and Images

We’re introducing a neural network called CLIP which efficiently learns visual concepts from natural language supervision. CLIP can be applied to any visual classification benchmark by simply providing the names of the visual categories to be recognized, similar to the “zero-shot” capabilities of GPT-2 and GPT-3.

Building an early warning system for LLM-aided biological threat creation

We’re developing a blueprint for evaluating the risk that a large language model (LLM) could aid someone in creating a biological threat. In an evaluation involving both biology experts and students, we found that GPT-4 provides at most a mild uplift in biological threat creation accuracy. While this uplift is not…

New and Improved Embedding Model for OpenAI

We are excited to announce a new embedding model which is significantly more capable, cost effective, and simpler to use.

GPT-5.1 for Developers

GPT-5.1 is now available in the API, bringing faster adaptive reasoning, extended prompt caching, improved coding performance, and new apply_patch and shell tools.

Advanced coding in low-resource programming languages

Learn how OpenAI is improving ChatGPT across low-resource programming languages, with new evaluations for software engineering, SQLite, and COBOL migration.

PaperBench

We introduce PaperBench, a benchmark evaluating the ability of AI agents to replicate state-of-the-art AI research.

Teen safety, freedom, and privacy

Explore OpenAI’s approach to balancing teen safety, freedom, and privacy in AI use.

OpenAI's comment to the NTIA on open model weights

OpenAI’s comment to the NTIA on open model weights This comment was submitted by OpenAI in response to NTIA’s March 2024 Request for Information on Dual-Use Foundation Models with Widely Available Weights.

Our framework for reporting model misalignment

OpenAI shares a framework for tracking, investigating, and disclosing model misalignment, alongside six reports of unexpected or concerning model behavior.

OpenAI releasing new open model in coming months, seeks community feedback

We’re planning to release our first open language model since GPT‑2 in the coming months. We’re excited to collaborate with developers, researchers, and the broader community to gather inputs and make this model as useful as possible.

Stargate Norway

We’re launching Stargate Norway—OpenAI’s first AI data center initiative in Europe under our OpenAI for Countries program. Stargate is OpenAI’s overarching infrastructure platform and is a critical part of our long-term vision to deliver the benefits of AI to everyone.

OpenAI and journalism

We support journalism, partner with news organizations, and believe The New York Times lawsuit is without merit.

DALL·E 3 is now available in ChatGPT Plus and Enterprise

We developed a safety mitigation stack to ready DALL·E 3 for wider release and are sharing updates on our provenance research.

OpenAI's response to the Axios developer tool compromise

OpenAI responds to the Axios supply chain attack by rotating macOS code signing certificates, updating apps, and confirming no user data was compromised.

GPT‑Rosalind for life sciences research

OpenAI introduces GPT-Rosalind, a frontier reasoning model built to accelerate drug discovery, genomics analysis, protein reasoning, and scientific research workflows.

Due to concerns about malicious applications, GPT2 will not be released (2019)

We’ve trained a large-scale unsupervised language model which generates coherent paragraphs of text, achieves state-of-the-art performance on many language modeling benchmarks, and performs rudimentary reading comprehension, machine translation, question answering, and summarization—all without task-specific training.

Aligning Language Models to Follow Instructions

We’ve trained language models that are much better at following user intentions than GPT-3 while also making them more truthful and less toxic, using techniques developed through our alignment research. These InstructGPT models, which are trained with humans in the loop, are now deployed as the default language models…

How the voices for ChatGPT were chosen

How the voices for ChatGPT were chosen We worked with industry-leading casting and directing professionals to narrow down over 400 submissions before selecting the 5 voices.

A letter from Sam & Jony

Building a family of AI products for everyone.

Scaling AI for everyone

Today we’re announcing $110B in new investment at a $730B pre money valuation. This includes $30B from SoftBank, $30B from NVIDIA, and $50B from Amazon.

WebGPT: Improving the factual accuracy of language models through web browsing

We’ve fine-tuned GPT-3 to more accurately answer open-ended questions using a text-based web browser.

Introducing GPT-5

We are introducing GPT‑5, our best AI system yet. GPT‑5 is a significant leap in intelligence over all our previous models, featuring state-of-the-art performance across coding, math, writing, health, visual perception, and more.

Aligning language models to follow instructions

We’ve trained language models that are much better at following user intentions than GPT-3 while also making them more truthful and less toxic, using techniques developed through our alignment research. These InstructGPT models, which are trained with humans in the loop, are now deployed as the default language models…

CLIP: Connecting text and images

We’re introducing a neural network called CLIP which efficiently learns visual concepts from natural language supervision. CLIP can be applied to any visual classification benchmark by simply providing the names of the visual categories to be recognized, similar to the “zero-shot” capabilities of GPT-2 and GPT-3.