AI Model Release Cycle Shortens to 44 Days in China and the United States

The average release cycle for new artificial intelligence models in China and the United States has been reduced to 44 days as of April 2026, a significant decrease from the previous average of 125 days. This finding is based on data from five American companies, including Anthropic and OpenAI, and four Chinese firms, such as Alibaba Group and Moon's Dark Side. The analysis covers the period from January 2023 to March 2026, highlighting a rapid acceleration in the development of AI technologies.

In September 2026, several companies have made notable advancements in their AI offerings. OpenAI and Anthropic released new models, while Meta has been updating its Muse Spark product monthly since July. Google introduced an updated version of its Gemini model on September 2, following a three-week interval. Additionally, Grok has launched new models for three consecutive months leading up to August. In China, DeepSeek has also been releasing updates monthly since July, with Alibaba and Z.AI following suit after August with their GLM series.

Anthropic reported that as of August 2026, 26% of its development work was driven by AI, a dramatic increase from nearly zero in February of the same year. OpenAI noted that its AI agents operated 3.1 times longer than human researchers in August, with average daily usage costs for ordinary researchers around $600, while the top 10% of users exceeded $7,000. Furthermore, the amount of code generated by OpenAI in August was seven times higher than the average output recorded in 2025.

© 2026 KLEA News. All Rights Reserved. This article is provided for informational purposes only. It is not offered or intended to be used as legal, tax, investment, financial, or other advice.

Source: KLEA News