Hunyuan T1 has the highest performance in the knowledge benchmark MMLU PRO and ... Chinese AI company DeepSeek releases 'DeepSeek-R1-Lite-Preview', an inference AI model comparable to OpenAI ...

Meta has launched Llama 4, its latest open-weight AI models, including Scout, Maverick, and Behemoth, offering advanced ...

Mark Zuckerberg’s Meta on Saturday launched its latest generation of Llama (large language model) artificial intelligence ...

Llama 4 Scout and Llama 4 Maverick—marking a significant step forward in its race against ChatGPT and Gemini. Both models are ...

Mark Zuckerberg's Meta on Saturday launched its latest generation of Llama (large language model) artificial intelligence ...

Chief Executive Officer Jensen Huang has argued that computation demand will grow even with the advent of more efficient models like DeepSeek’s R1 ... Lite model did better in a key benchmark ...

Meta's new AI models aren’t just about text — they’ve been built to understand and reason through images and videos as well.

Pro benchmark, a test that thoroughly assesses large language models' language comprehension and reasoning across diverse domains. The score beat DeepSeek-R1’s 84 points but trailed the 89.3 ...

Pro benchmark, a test that thoroughly assesses large language models' language comprehension and reasoning across diverse domains. The score beat DeepSeek-R1’s 84 points but trailed the 89.3 ...

Also Read: Two AI models pass benchmark ... DeepSeek toppling heavyweights like OpenAI. Also Read: Meta’s AI research head to step down amid company's $65 billion computing push After DeepSeek's ...

Tencent had previously released a preview ... R1 model. A chart published in the post comparing the T1 model to DeepSeek R1 showed Tencent's outperformed on some knowledge and reasoning benchmarks.