Unverified50% confidenceFactExact time
The longer the conversation, the more expensive it gets toward the end, because every time a sentence is sent, the model reads all the preceding content again from the start, with the parts already read counted as cache hits (lower price) and the parts not yet read charged at the original price.
1
Sources
50%
Confidence
Long-term
Relevance
11/6/2026
First Seen
Sources
4 Must-Have Plugins for Saving Tokens in DSH — Stop Installing Blindly
bilibili天才依山尽10/29/2026
Related Entities
Related Claims
UnverifiedFor multi-agent workflows that repeatedly read long contexts, the cached input discount can make the actual cost far lower than the sticker price, making it a key cost item that is easily overlooked when building products.75% similarUnverifiedSwitching models mid-session breaks the cache, causing the entire conversation to be reprocessed at full price, making it one of the most expensive operations.74% similarUnverifiedIncluding large amounts of conversation history or background documents (long context) in each request causes costs to accumulate in a non-linear manner.74% similarVerifiedToken billing is calculated based on the total volume of input and output; the longer a conversation, the more each call pays for an increasingly bloated history.72% similarUnverifiedFor agent or chat products that resend long system prompts and history on every turn, cache prices may have a greater impact on the bill than input prices.71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1142145API
curl https://kongchang.com/api/v1/knowledge/claims/1142145MCP
get_claim(id=1142145)