Aliyun's large model service platform, Bailian, has issued a notice stating that starting from Beijing Time November 13, 2025, it will implement a price reduction for the Tongyi Qianwen 3-Max model on the China Station - Beijing. After the price reduction, the batch invocation cost will be halved; for implicit caching, the portion that hits the cache will be billed at 20% of the standard unit price per input token; for explicit caching, the tokens used to create the cache will be billed at 125% of the standard unit price per input token, with subsequent hits only requiring a 10% payment.
