4/16/2025
Thinking with images
OpenAI o3 and o4-mini models achieve visual reasoning breakthrough by incorporating images in chain-of-thought processing.
4/15/2025
Generate videos in Gemini and Whisk with Veo 2
Video generation in Gemini and Whisk using Veo 2. Feature announcement for generative models.
4/14/2025
Introducing GPT-4.1 in the API
OpenAI releases GPT-4.1 API with improvements in coding, instruction following, and long-context understanding, including first nano model.
4/10/2025
BrowseComp: a benchmark for browsing agents
BrowseComp is a benchmark for evaluating and comparing browsing agents' ability to navigate and retrieve information from web environments.
4/2/2025
PaperBench: Evaluating AI’s Ability to Replicate AI Research
PaperBench is a benchmark for evaluating AI agents' ability to replicate state-of-the-art AI research end-to-end.
3/27/2025
Moving from intent-based bots to proactive AI agents
Zendesk pilots new AI agents powered by OpenAI models that autonomously plan and execute customer service responses beyond intent-based conversation management.
3/26/2025
Security on the path to AGI
OpenAI shares updates on Cybersecurity Grant Program, bug bounties, and security initiatives including research on prompt injection and AGI safety.
3/25/2025
Automating 90% of finance and legal work with agents
Hebbia's Matrix platform uses multiple AI agents in parallel to automate complex financial and legal workflows, orchestrating OpenAI's o3-mini and o1 models.
3/25/2025
Scaling the OpenAI Academy
OpenAI Academy expands with free online resource hub for AI literacy, best practices, and tools for diverse audiences.
3/21/2025
Early methods for studying affective use and emotional well-being on ChatGPT
OpenAI and MIT Media Lab research collaboration studying affective use and emotional well-being impacts of ChatGPT interactions.
3/20/2025
Personalizing travel at scale with OpenAI
Booking.com integrates OpenAI LLMs to deliver smarter search, faster support, and intent-driven travel experiences at scale.
3/18/2025
New in ChatGPT for Business: March 2025
ChatGPT for Business March 2025 updates focus on interactive, customizable, and agentic capabilities for teams.
3/12/2025
Experiment with Gemini 2.0 Flash native image generation
Native image generation in Gemini 2.0 Flash for developers. Feature addition for developer use.
3/11/2025
New tools for building agents
OpenAI releases first building blocks for developers and enterprises to build useful and reliable AI agents with complex multi-step task capabilities.
3/7/2025
Nubank elevates customer experiences with OpenAI
Nubank leverages OpenAI solutions to enhance customer experience and internal efficiency for 114M users across Latin America.
3/6/2025
Accelerating engineering cycles 20% with OpenAI
Factory Platform uses OpenAI o1, o3-mini, and GPT-4o for software development workflows, achieving 20% engineering cycle acceleration.
3/4/2025
LaunchDarkly's approach to AI-powered product management
LaunchDarkly CPTO discusses AI-powered product management and the evolving role of product managers in AI applications.
2/27/2025
Supporting sellers with enhanced product listings
Mercari marketplace uses GPT-4o mini and GPT-4 for product listing enhancement and seller support features.
2/27/2025
OpenAI GPT-4.5 System Card
OpenAI GPT-4.5 system card detailing risk assessment, preparedness scorecard, and deployment criteria for the new model.
2/27/2025
Introducing GPT-4.5
OpenAI releases GPT-4.5, largest model yet with improved pattern recognition and natural interaction, available to Pro users and developers.
2/25/2025
Deep research System Card
OpenAI Deep Research System Card detailing safety evaluations, red teaming, and risk mitigations before release.
2/25/2025
Estonia and OpenAI to bring ChatGPT to schools nationwide
Estonia and OpenAI partnership to provide ChatGPT Edu access to secondary school students and teachers.
2/20/2025
College students and ChatGPT adoption in the US
Survey of ChatGPT adoption among US college students and implications for workforce readiness.
2/18/2025
Introducing the SWE-Lancer benchmark
SWE-Lancer benchmark evaluates frontier LLMs on real-world freelance software engineering tasks and earnings potential.
2/5/2025
Deep Dive into LLMs like ChatGPT
General overview of large language model architectures and characteristics similar to ChatGPT.
1/31/2025
OpenAI o3-mini System Card
Safety evaluation report for OpenAI o3-mini model including red teaming and preparedness framework assessments.
1/28/2025
Introducing ChatGPT Gov
ChatGPT Government edition for U.S. government agencies to access OpenAI models.
1/23/2025
Operator System Card
OpenAI safety documentation for AI systems covering prompt injection, jailbreak, privacy, and red teaming mitigations.
1/23/2025
Computer-Using Agent
Universal interface enabling AI agents to interact with digital environments and applications.
1/22/2025
Trading inference-time compute for adversarial robustness
Research on improving adversarial robustness by trading inference-time compute in language models.
12/20/2024
Deliberative alignment: reasoning enables safer language models
New alignment strategy for o1 models using reasoning to teach safety specifications.
12/17/2024
FACTS Grounding: A new benchmark for evaluating the factuality of large language models
Benchmark measuring LLM factuality and hallucination reduction. Evaluation tool for LLM reliability.
12/11/2024
Introducing Gemini 2.0: our new AI model for the agentic era
Gemini 2.0 multimodal model positioned for agentic applications. Core LLM release for agent-centric era.
12/11/2024
Boosting the customer retail experience with GPT-4o mini
Zalando uses GPT-4o mini to power customer service assistant improving retail experience.
12/9/2024
Sora System Card
OpenAI's Sora video generation model technical overview. Generates video from text/image/video inputs up to 1080p.
12/5/2024
OpenAI o1 System Card
OpenAI o1 safety report detailing red teaming, risk evaluations, and preparedness framework.
11/21/2024
Advancing red teaming with people and AI
Red teaming methodology combining human and AI-based safety testing for model evaluation.
11/20/2024
Building smarter maps with GPT-4o vision fine-tuning
Fine-tuning GPT-4o vision for improved map building and spatial understanding.
10/31/2024
Introducing ChatGPT search
ChatGPT search feature provides web-sourced answers with citations.
10/31/2024
Promega’s top-down adoption of ChatGPT accelerates manufacturing, sales, and marketing
Case study: Promega accelerates manufacturing and sales operations using ChatGPT.
10/23/2024
Simplifying, stabilizing, and scaling continuous-time consistency models
Simplified continuous-time consistency models achieving comparable quality to diffusion models with only two sampling steps.
10/17/2024
Solving complex problems with OpenAI o1 models
OpenAI o1 reasoning models applied to coding, strategy, and research problems. Video overview of capabilities.
10/15/2024
Evaluating fairness in ChatGPT
Research analyzing ChatGPT response fairness based on user names using privacy-preserving AI assistants.
10/10/2024
MLE-bench: Evaluating Machine Learning Agents on Machine Learning Engineering
MLE-bench: Benchmark for evaluating AI agents on machine learning engineering tasks.
10/1/2024
Introducing vision to the fine-tuning API
OpenAI announces fine-tuning API now supports vision models (GPT-4o) for improved image and text capabilities.
10/1/2024
Prompt Caching in the API
OpenAI API feature providing automatic cost discounts for repeated input tokens through prompt caching.
10/1/2024
Creating agent and human collaboration with GPT 4o
Altera uses GPT-4o to enable human-AI collaboration in their platform.
9/26/2024
Upgrading the Moderation API with our new multimodal moderation model
OpenAI releases multimodal moderation model based on GPT-4o for detecting harmful text and images in developer applications.
9/24/2024
Updated production-ready Gemini models, reduced 1.5 Pro pricing, increased rate limits, and more
Google releases updated production-ready Gemini models with reduced pricing and increased rate limits.
9/24/2024
Introducing Verdi, an AI dev platform powered by GPT-4o
Mercado Libre introduces Verdi, an AI developer platform powered by GPT-4o.