Baidu, the Chinese tech giant, on Friday unveiled its latest AI models, ERNIE X1 Turbo and ERNIE 4.5 Turbo. These AI models are faster, more innovative, and more cost-effective, building on the successes of their predecessors.
These improved versions are intended to provide improved performance across multimodal understanding, logical reasoning, and tool usage, catering to developers, researchers, and companies alike.
The launch occurred in China, with global availability expected soon through Baidu’s Qianfan platform and other cloud-based services. These Turbo models, which aim to make high-level AI more accessible, are also incredibly cost-efficient—ERNIE X1 Turbo costs only $0.14 per 1 million input tokens and $0.55 per 1 million output tokens, which is roughly 75% less than DeepSeek R1, according to Baidu.
These enhancements are made possible by advancements in chain-of-thought reasoning, enhanced multimodal capabilities, and improved tool integration, positioning ERNIE X1 Turbo and 4.5 Turbo in strong contention for the most efficient and intelligent AI systems on the market today.
Before launching its latest AI model, Baidu had introduced some of its models earlier in March, including ERNIE 4.5, a native multimodal model, and ERNIE X1, a deep-thinking reasoning model. The company has also made its AI chatbot, ERNIE Bot, free for individual users earlier than planned.
Notably, Baidu was among the first in China to launch a Chatgpt-style chatbot in early 2023, with Ernie 4.0 claiming to rival OpenAI’s GPT-4. However, adoption has lagged due to strong competition, especially from DeepSeek’s R1.
[Note: An earlier version of the article stated that the ERNIE X1 Turbo is 25% cheaper. This has now been corrected.]
