Tencent Cloud's Hunyuan Large Model Updates and Pricing Adjustments
TapTechNews May 30th news, Liu YuHong, the vice president of Tencent Cloud and the person in charge of Tencent's Hunyuan large model, stated that there are two main reasons for the current low penetration rate of large models: on one hand, the large model is still in the development stage, and the AI product solutions are not mature; on the other hand, the user perception is still being established, and the demand for AI is mainly focused on search/efficiency.
Liu YuHong said that the original search engine may have already done very well, and now the large model can make the search better. Since Tencent released the large model product last September, the model efficiency has increased by 50%, from thousands of parameters to trillions of parameters, supporting 256k native window context, and also supporting 16-second video generation. Currently, more than 600 internal businesses of Tencent have accessed the Hunyuan large model.
On May 22nd, Tencent Cloud announced a new large model pricing plan. Among them, one of the main models, Hunyuan-lite model, plans to upgrade the total length of API input and output from the current 4k to 256k, and the price is adjusted from $0.008 per thousand tokens to completely free.
Hunyuan-standard API input price is reduced from $0.01 per thousand tokens to $0.0045 per thousand tokens, a decrease of 55%, and the API output price is reduced from $0.01 per thousand tokens to $0.005 per thousand tokens, a decrease of 50%. The newly launched Hunyuan-standard-256k has the ability to handle ultra-long text of more than 380,000 characters. The API input price is lowered to $0.015 per thousand tokens, a decrease of 87.5%, and the API output price is lowered to $0.06 per thousand tokens, a decrease of 50%.
The highest configuration trillion-parameter model Hunyuan-pro, the API input price is reduced from $0.1 per thousand tokens to $0.03 per thousand tokens, a reduction of 70%.
The above price adjustments take effect immediately. At the same time, in order to better serve developers, the one-stop intelligent agent creation and distribution platform, Tencent Yuanqi, has also comprehensively upgraded the model resource support plan: users can not only distribute their dedicated AI intelligent agent to all channels of Tencent with one click, obtain free Hunyuan large model resources and traffic support, but also distribute and deploy the intelligent agent in the form of API to various scenarios, and the free token quota has been increased from the previous 1 million to 100 million.