DeepSeek V4-Flash API has opened public beta. Its speed and cost-effectiveness comprehensively crush GPT. Recently, DeepSeek officially announced that its latest version of API has opened public beta.

DeepSeek V4-Flash API has launched public beta, and its speed and cost-effectiveness have completely surpassed GPT. Recently, DeepSeek officially announced that its latest version of API has launched public beta. As a high-profile AI product, the new version has achieved a disruptive upgrade in agent capabilities, and its benchmark test results have even surpassed previous preview versions, demonstrating extremely strong overall strength. In actual application testing, some deep AI developers immediately connected the new version of the API to the system for actual testing. The results showed that when faced with the same difficult and complex tasks and prompt words, the new version only took about 40 seconds to complete efficiently, while the top overseas large model of the same type took 1 minute and 47 seconds. Although overseas models still have a slight advantage in some refined output quality, the delivery speed and usability of the new version have fully reached the high level of the industry. In addition to the amazing processing speed, price and cost have also become the core highlights of this upgrade. According to the latest official pricing mechanism, the input cost of some versions has dropped significantly, and the overall decline has been significant. However, different versions of the strategy have different priorities, and the API costs for some high-performance or specific periods of time have been adjusted. Overall, compared with overseas mainstream official large models, domestic AI still maintains a highly competitive advantage in API pricing, with the cost gap generally ranging from dozens to hundreds of times. Industry insiders believe that as domestic large-scale models continue to make breakthroughs in performance, speed and cost control, the threshold for commercialization of AI technology is further lowering. In the future, more developers and enterprises will be able to enjoy efficient AI computing power support at extremely low costs, and the industry ecosystem may usher in a new round of reshaping. via AI News (author: AI Base)