GuShiio.com Article source: Technology whirlpool
Image source: Generated by AI
Thanks to DeepSeek, all premium models are now free.
Recently, Wenxinyiyan announced on the official website that it will be fully free from 0:00 on April 1. All PC and APP users can experience the latest model of the Wenxin series. OpenAI CEO Sam Altman also announced that GPT-4.5/5 will be released soon, and the free version of ChatGPT will use GPT-5 for conversations without restrictions under standard smart settings.
Not long after the Hundred Models War, now they are starting to open books on the road to free. Why?
DeepSeek stirs up the market? Wenxin announced that it was free
On February 13, Wenxinyiyan announced on the official website that with the iterative upgrade of the Wenxinyiyan model and the continuous decline in costs, Wenxinyiyan will be fully free from 0:00 on April 1, and all PC and APP users can experience the latest model of the Wenxinyan series.
Wenxinyiyan’s official website also simultaneously revealed that the in-depth search function will be launched from now on, which can provide users with expert-level content responses and handle multi-scene tasks to achieve multimodal input and output.
There may be two reasons why Baidu Wenxin’s words are free. The first is that the cost of iteration and reasoning for large models is declining, and the other is to cope with DeepSeek’s outstanding performance.
At the World Governments Summit 2025 Summit held in Dubai, United Arab Emirates, Baidu founder Robin Li said that when we talk about Moore’s Law, performance will double and costs will be halved every 18 months; but today, when we talk about large language models, it can be said that every 12 months, reasoning costs can be reduced by more than 90%.
When costs drop, the corresponding price will inevitably decrease accordingly, so as to maintain competitiveness in the market. From the Spring Festival to the present, the popular DeepSeek has not cooled down due to the passage of time. Instead, it has become more and more popular due to the continuous access of mobile phone manufacturers and car companies, which has also made many large models feel the crisis of popularity.
Huawei was the first to announce that the Xiaoyi Assistant App will be connected to DeepSeek-R1. Immediately afterwards, it proudly claimed that the YOYO smartphone store has launched the initial experience version of DeepSeek-R1. OPPO officials announced that its latest flagship Find N5 will be directly connected to DeepSeek-R1. Meizu confirmed that its mobile phone products and AI ecosystem products will gradually integrate DeepSeek technology. Nubia announced that it will deeply embed a full-blooded version of DeepSeek-R1.
In terms of car companies, more than 20 car companies and brands have announced access to DeepSeek. Major car companies such as BYD, Geely, Chery, Chang ‘an, SAIC, Dongfeng, and GAC are all included. Of course, new car companies or brands such as Lantu, Zhiji, and Jikrypton, as well as car service providers such as Zebra Zhixing and Yikatong, have also joined the cooperation ecosystem.
In fact, many mobile phone manufacturers and car companies have also connected to the Wenxin Model API before, so Baidu Wenxin Yiyan announced that it will be free at this time. On the one hand, it is due to the plan and on the other hand, it may want to regain some of the popularity. Let people pay more attention to Wenxin Yiyan.
Stimulated by DeepSeek, OpenAI is full of fire
DeepSeek’s popularity has changed OpenAI’s past limpness and begun a series of measures to try to prove its dominance in the field of large models.
In a podcast broadcast recently, OpenAI CEO Ultraman said that in order to prevent OpenAl from being surpassed, there is pressure every day, and revealed that he would communicate and meet with the Deepseek team.
He said that if you want to avoid being surpassed, you have to wake up every day and be vigilant. I hope OpenAI will not repeat the mistakes of Napster (the first P2P music sharing platform that ended in failure due to commercial pressure), because we have always attached great importance to this issue.
At 3 a.m. Beijing time on February 13, OpenAI CEO Sam Altman announced that GPT-4.5/5 will be released soon. The free version of ChatGPT will use GPT-5 for conversations without restriction under standard smart settings.
It is reported that OpenAI soon released GPT-4.5, which is the legendary Orion model and the last non-thought chain model. Then we will integrate the two major models of GPT and O series to create a new system that can automatically select thinking and non-thinking functions and is suitable for multiple tasks.
Sam Altman said that in ChatGPT and APIs, OpenAI will release GPT-5 as a system that integrates many technologies, including the o3 model. Subsequently, the o3 model was no longer launched as an independent model. Plus subscribers will be able to run GPT-5 with a higher level of intelligence, and Pro subscribers will be able to run GPT-5 with a higher level of intelligence. These models will integrate more functions such as voice, graphics, search, and in-depth research.
But this is obviously a preview. Regarding how long GPT-4.5 and GPT-5 will take to arrive, Sam Altman responded that it will be months or weeks.
In addition to the news of the OpenAI model, it also spent a lot of money on publicity. Previously, OpenAI made its debut in the Super Bowl (the National Football League’s annual championship game) with a 60-second ad that cost approximately $14 million in the first half of the year. This series of measures can be seen that OpenAI attaches great importance to the rise of DeepSeek, and Google, also a technology giant, cannot sit still.
Faced with the trend of low prices and efficiency set off by DeepSeek, in early February, Google officially launched the flagship AI model Gemini 2.0 Pro Experimental, and also released the Gemini 2.0 Flash Thinking model. It is regarded as an important move for Google to actively respond to competition in the AI field and consolidate its market position.
written in the end
The appearance of DeepSeek is like a stone thrown into the center of a lake, arousing thousands of waves in the “lake” of the large model. Its free strategy and efficient performance have not only redefined the competitive landscape of the industry, but also forced other giants to speed up their pace and adjust their strategies to cope with the impact of this strong opponent.
In the future, competition in the field of large models may become more intense, but this healthy competition will also promote the emergence of more high-quality products and services, and the ones who will ultimately benefit will be global users.
