eWEEK content and product recommendations are editorially independent. We may make money when you click on links to our partners. Learn More.
Chinese AI company DeepSeek has disrupted the market by offering its advanced AI model for free, forcing competitors to rethink their pricing strategies. In response, Baidu and OpenAI have announced plans to make their next-generation AI models available to non-subscribers.
Baidu chatbot is a Chinese rival to DeepSeek
Premium features of Baidu’s chatbot will be available to all users for free starting April 1, the company announced on Feb. 13. Premium features include deep search and multimodal input and output. Baidu followed the announcement with another on Friday, saying the next generation of its Ernie AI model will be open source as of June 30.
Previously, premium features on Ernie started at 59.9 yuan ($8.20) per month.
DeepSeek surprised AI makers worldwide by announcing a deep reasoning model that performed comparably to OpenAI’s advanced models on certain benchmark tests but cost much less to train. Users can try DeepSeek’s advanced R1 reasoning model for free. However, security researchers have found critical flaws and have concerns about possible ties to the Chinese government.
GPT-5 branding changes may be in response to DeepSeek’s free reasoning
U.S. tech darling OpenAI announced on Feb. 12 that its next model, GPT-5, will be free to all users of ChatGPT when it comes out at an unspecified time in the next few months. Currently, the basic version of ChatGPT, including image generation, is free. A Pro subscription, which adds unlimited access to the more powerful models, costs $200 per month. After GPT-5 is released, Pro subscribers will gain access to “an even higher level of intelligence.”
OpenAI CEO Sam Altman did not focus on price in his announcement of GPT-5. He said OpenAI wants to “do a better job of sharing our intended roadmap, and a much better job simplifying our product offerings.”
That means ChatGPT users will no longer see an extensive model picker displaying GPT-family models and o-family models. Instead, “a top goal for us is to unify o-series models and GPT-series models by creating systems that can use all our tools, know when to think for a long time or not, and generally be useful for a very wide range of tasks.” As generative AI infrastructure becomes more efficient, companies may be able to offer more products for free or as relatively inexpensive subscriptions. DeepSeek’s disruptive move could mark a pivotal moment in the industry’s approach to monetization.