Introducing SWE-bench Verified

We’re releasing a human-validated subset of SWE-bench that more reliably evaluates AI models’ ability to solve real-world software issues.

Introduction to ggml

Tool Use, Unified

Welcome Falcon Mamba: The first strong attention-free 7B model

Zico Kolter Joins OpenAI’s Board of Directors

Zico Kolter Joins OpenAI’s Board of Directors We’re strengthening our governance with expertise in AI safety and alignment. Zico will also join the Safety & Security Committee

Enabling a data-driven workforce

In this video, we share practical examples of how employees can use ChatGPT Enterprise to efficiently analyze data and uncover insights.

XetHub is joining Hugging Face!

GPT-4o System Card

Pairing data with APIs to unlock customer value

Rakuten Pairs Data with AI to Unlock Customer Insights and Value

Introducing Structured Outputs in the API

We are introducing Structured Outputs in the API—model outputs now reliably adhere to developer-supplied JSON Schemas.

Introducing TextImage Augmentation for Document Images

2024 Security Feature Highlights

Google releases Gemma 2 2B, ShieldGemma and Gemma Scope

Memory-efficient Diffusion Transformers with Quanto and Diffusers

Serverless Inference with Hugging Face and NVIDIA NIM

LAVE: Zero-shot VQA Evaluation on Docmatix with LLMs - Do We Still Need Fine-Tuning?

SearchGPT is a prototype of new AI search features

We’re testing SearchGPT, a temporary prototype of new search features that give you fast and timely answers with clear and relevant sources.

Improving Model Safety Behavior with Rule-Based Rewards

We’ve developed and applied a new method leveraging Rule-Based Rewards (RBRs) that aligns models to behave safely without extensive human data collection.

Llama 3.1 - 405B, 70B & 8B with multilinguality and long context

WWDC 24: Running Mistral 7B with Core ML