By combining the efficiency of a Mixture-of-Experts architecture with the openness of an Apache 2.0 license, OpenAI is providing a way for many enterprises to more easily, cheaply and safely redact PII data.

Chinese artificial intelligence developer DeepSeek today released a new series of open-source large language models. V4, as the algorithm family is called, comprises two LLMs on launch. There’s the flagship V4-Pro and a smaller model called V4-Flash that trades off some output quality for lower hardware usage.

Recent benchmark results show open-weight AI models and OpenAI’s GPT-5.5 surpassing established systems like Claude and Gemini in complex task performance. GPT-5.5 scored 82.7% on Terminal-Bench 2.0, while Chinese-developed GLM-5.1 and Kimi K2.6 ...

On March 31, OpenAI cofounder and CEO Sam Altman announced on X that his company is working on a “powerful new open-weight language model,” and that it would be released in the coming months. It’s a notable turn for OpenAI, which hasn’t released an ...

Discover how DeepSeek 4 rivals closed-source AI in 2026 with open weights, reduced FLOPs, and advanced hardware validation on Havi Ascent NPUs.

Chinese-developed open-weight AI models GLM-5.1 and Kimi K2.6 have surpassed leading closed-source systems from OpenAI, Google, and Anthropic in the SWE-Bench Pro coding benchmark. Both models, released this month, are available under permissive licenses ...

OpenAI just dropped two new open-weights models. Here's why that isn't the same as being fully open. Katelyn is a reporter with CNET covering artificial intelligence, including chatbots, image and video generators. Her work explores how new AI technology ...

HONG KONG, CHINA - JANUARY 28: In this photo illustration, the DeepSeek logo is seen on a phone in front of a flag of China on January 28, 2025 in Hong Kong, China. American investors are confronting an uncomfortable reality in artificial intelligence: the ...