OpenAI o3 and o4-mini: Tool-Using Reasoning Models
OpenAI released o3 and o4-mini reasoning models that can use web browsing, code execution, and file analysis tools during the thinking process.
OpenAI's 12 Days of Shipmas: o1, Sora, Canvas, Projects, and More
OpenAI ran a '12 Days of Shipmas' campaign in December, announcing a new product each day. o1 full model, Sora public release, and ChatGPT Pro were major highlights.
OpenAI o1: The AI Model That Thinks Before Answering
OpenAI's new o1 model introduces the "chain-of-thought reasoning" paradigm. It achieves PhD-level results in math and science by thinking step-by-step.
GPT-4o: OpenAI's Omni Model Processes Text, Audio, and Vision Together
OpenAI's new model GPT-4o can natively process text, audio, and visual inputs simultaneously. Response time has dropped to 232 milliseconds.
OpenAI Unveils Sora: Realistic Video Generation from Text Is Now Possible
OpenAI's new text-to-video model Sora can generate high-quality videos up to 60 seconds from a short text description. The results shocked the industry.