Official release from AI companies

xAI has launched Grok 3, outperforming Grok 2 by 9% in data extraction accuracy. It excels in complex legal tasks, multi-step reasoning, and quantitative analysis, while still needing improvements in language precision.

Grok 3 has been launched by xAI, showcasing exceptional performance in complex enterprise tasks across sectors like finance, healthcare, and legal, with users reporting unique capabilities not found in other models.

xAI introduces the Grok 3 family, featuring Grok 3 Mini, which outperforms reasoning models at a fraction of the cost. It excels in STEM, math, and coding, offering unprecedented affordability and access to detailed reasoning traces.

xAI introduces Grok 3 Mini, a cost-effective AI model that excels in graduate-level STEM, math, and coding tasks, outperforming pricier competitors. It offers full reasoning traces in API responses, enhancing transparency.

xAI introduces the Grok 3 family on their API, featuring Grok 3 Mini, which outperforms reasoning models at a significantly lower cost, and Grok 3, the strongest non-reasoning model, excelling in law, finance, and healthcare tasks.

OpenAI has launched a new library for ChatGPT image creations, allowing users to access all their images in one place and share them directly from the platform.

Google DeepMind has launched Gemini 2.5 Flash, a hybrid reasoning model that allows developers to adjust its reasoning capacity based on task complexity, optimizing for cost and speed in applications like chat apps and data extraction.

Google DeepMind announces the release of Gemini 2.5 Flash, a hybrid reasoning model that allows users to control its cognitive capacity based on budget, making it suitable for various applications like chat apps and data extraction.

Meta FAIR has unveiled several new research artifacts aimed at advancing machine intelligence, including the Meta Perception Encoder, a vision-language model, and the Dynamic Byte Latent Transformer, which redefines language model efficiency.

OpenAI has launched models o3 and o4-mini, which integrate uploaded images into their processing, allowing them to think with images. These models are now available for ChatGPT Plus, Pro, and Team users.

OpenAI has introduced new models, o3, o4-mini, and o4-mini-high, for ChatGPT Plus, Pro, and Team users, with Enterprise and Edu access coming soon. Developers can also use these models via the Chat Completions API.

OpenAI has launched its latest models, o3 and o4-mini, which enhance reasoning capabilities by integrating tools like web search and image analysis. o3 excels in various domains, while o4-mini offers speed and cost-efficiency for high-volume tasks.