Stacy Pereira
- Tech giant Baidu released its flagship generative AI model Ernie 4.5 and model X1, to the open-source community.
- The move will pressurize companies like OpenAI, Anthropic to justify gated APIs and premium pricing.
- Baidu’s open-source strategy could be a step toward democratizing advanced AI.
China’s tech giant Baidu announced that from June 30, 2025, it will open‑source its flagship generative AI model, ‘Ernie Bot’, making it freely available to developers and businesses worldwide. This marks a dramatic shift from Baidu’s previous reliance on proprietary systems, aiming to foster a robust developer ecosystem and challenge global AI powerhouses. Many are viewing this as China’s 2nd Deep Seek moment.
By releasing Ernie’s code, Baidu is intentionally converting high-performance AI into a commodity. Back in February, Baidu had confirmed that Ernie would be made available to the open community, however it wasn’t known then whether the AI models would be fully open-sourced or partially. Analysts suggest this move is designed to undercut license fees from Western competitors like OpenAI and Anthropic.
Pricing pressure for competitors
Newsletter
Get weekly updates on the newest crypto stories, case studies and tips right in your mailbox.
The move will ignite a new wave of price competition, putting pressure on closed providers like OpenAI and Anthropic to justify gated APIs and premium pricing. As for domestic rivals, Baidu now matches Deep Seek’s open‑friendly approach, which will also catapult the likes of Alibaba, Tencent, ByteDance, and others to advance their own open‑source LLM.
According to a report from Business Insider, Ernie 4.5 and its reasoning counterpart Ernie X1 have been praised for delivering GPT‑4 level performance at a fraction of the cost. X1 reportedly offers parity with DeepSeek’s R1 at half the price, while Ernie 4.5 beats GPT‑4.5 benchmarks at just 1% of the cost.
Implications of open sourcing
Open‑sourcing Ernie is likely to reshape the global AI landscape. This would mean that academic researchers, startups, and smaller firms can now build on this world‑class LLM tech. Community contributions could rapidly improve the model’s performance and safety.
Experts say this move could help make AI more accessible worldwide. Sean Ren, a computer science professor at the University of Southern California and Samsung’s AI Researcher of the Year, told CNBC, “This goes beyond China. Whenever a major lab releases a strong model as open source, it pushes the whole industry forward.”
