Alibaba Cloud: Partial historical models of the large model service platform Bailian will be offline on July 2nd

On April 3rd, according to the notice on Alibaba Cloud's official website, some historical models of the large model service platform Bailian will be taken offline on July 2nd, 2025. From the date of the announcement, Alibaba Cloud will gradually lower the model traffic limit for these models until the models are officially taken offline. The traffic limit includes the number of calls per minute (QPM) and the number of tokens consumed per minute (TPM). Models that have applied for expansion will restore their default traffic limit. After the model is taken offline, it will no longer support model inference, some models will stop model fine-tuning, and applications and services that are calling the model will be unable to return results normally. Alibaba Cloud reminds that there is no direct substitute model for these models.