Baidu, a Chinese multinational technology company specializing in Internet services and Artificial Intelligence, plans to release the next generation of its AI model in the second half of this year, according to a source familiar with the matter.
The AI model known as “Ernie 5.0”, is expected to bring major improvements in multimodal capabilities, allowing it to process and generate text, video, images, and audio seamlessly.
Baidu’s plan to roll out its Artificial intelligence model comes following the launch of Chinese startup DeepSeek, which made headlines earlier this year, shaking the AI ecosystem. Recall that the Launch of DeepSeek’s latest AI model on 20 January, shook the Artificial intelligence universe.
The Hangzhou-based company released DeepSeek-R1, a partly open-source ‘reasoning’ model that can solve some scientific problems at a similar standard to o1, OpenAI’s most advanced LLM. DeepSeek’s AI Assistant, powered by DeepSeek-V3, reportedly overtook rival ChatGPT to become the top-rated free application available on Apple’s App Store in the United States. This raised doubts about the reasoning behind some U.S. tech companies’ decision to pledge billions of dollars in AI investment and shares of several big tech players, including Nvidia, were seriously hit.
DeepSeek claims to have spent around $5.5 million to train its V3 model, a considerably frugal approach to delivering the same results, that took the likes of Google, OpenAI, Meta, and others, hundreds of millions of dollars in investments to achieve. The startup’s innovative techniques, cost-efficient solutions, and optimization strategies have no doubt had an undeniable effect on the AI landscape, primarily because it challenges the high-cost barrier associated with building large language models (LLMs).
Notably, it has indirectly put pressure on AI Giants to justify high costs. Companies like OpenA1, Google, and Meta have justified their huge funding rounds by emphasizing the enormous compute costs required to train frontier models. If DeepSeek proves that cutting-edge Al can be built on a small budget, investors may push for more efficiency rather than just throwing money at bigger GPU clusters.
Speaking at the recent World Governments Summit in Dubai, Baidu CEO Robin Li highlighted rapid cost reductions in Al development, stating that the inference cost of foundation models has dropped by more than 90% in the past year. “If you can reduce the cost by a certain percentage, then that means your productivity increases by that kind of percentage. That’s the nature of innovation,” Li said.
It is worth noting that Baidu’s proposed roll-out of the AI model is coming after China’s e-commerce giant company Alibaba, last month, released a new version of its Qwen 2.5 artificial intelligence model that it claimed surpassed the highly-acclaimed DeepSeek-V3.
“Qwen 2.5-Max outperforms almost across the board GPT-4o, DeepSeek-V3 and Llama-3.1-405B,” Alibaba’s cloud unit said in an announcement posted on its official WeChat account, referring to OpenAI and Meta’s most advanced open-source AI models.
Amidst the roll-out of generative AI models in China, the country’s Al ecosystem is rapidly strengthening as Baidu, DeepSeek, and Alibaba, ramp up their Al capabilities to compete with U.S. models like OpenAl’s GPT-4 and Google’s Gemini. With DeepSeek’s cost-efficient model training making waves, Chinese tech giants are pushing the boundaries of large language models (LLMs), multimodal Al, and generative Al.
This surge in Al development signals a growing China-U.S. Al rivalry, with Chinese companies aiming to close the innovation gap and expand global Al influence.
