Skip to main content

extended tasks

Some models can continue to perform extended operations on the original task. These operations usually require telling us the original task ID in the submitted request.

cache

Many model manufacturers have prompt word caching capabilities, which can significantly reduce processing time and cost for repetitive tasks or prompt words with consistent elements. The caching feature generally cannot be used across accounts. As a transit cloud service platform, this platform cannot fully guarantee that multiple requests will go through the same channel.ClaudeThe model may be cached repeatedly as a result. The user can pass in the ID of the previous conversation so that new requests continue to use the channel of the previous conversation.
HeaderValueSourceValidity
X-Request-Id202511162255194339355347BE3UyUYResponse Header5 minutes
X-Response-Idmsg_013Ova2CMHLNnXjNJJKqJ9EFConversation ID in Response JSON5 minutes