Dhaka, March 16 -- Chinese tech giant Baidu has introduced two new artificial intelligence (AI) models, positioning them as superior to those of DeepSeek and OpenAI based on certain benchmarks, amid the intensifying competition in large language models (LLMs).

On Sunday, Baidu released its multimodal foundational model, Ernie 4.5, and its first multimodal reasoning model, Ernie X1, making them freely available on its website. According to Baidu, Ernie 4.5 outperformed OpenAI's GPT-4o in several benchmark tests, including CCBench and OCRBench, with its multimodal capabilities covering images, audio, and video.

The text performance of Ernie 4.5 also surpassed DeepSeek V3 on multiple benchmarks, while being comparable to OpenAI's GPT-4.5....