Unverified50% confidenceFactExact time
94% of the tokens consumed by the project were cached tokens, and cached tokens are typically 75%–90% cheaper than fresh tokens.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/16/2026
First Seen
Valid until: 12/15/2026
Sources
Related Entities
Related Claims
UnverifiedThe 90% token saving is the result of a specific scenario, and the actual saving depends on codebase size, usage habits, and task type.68% similarUnverifiedThe user's workload has a KV Cache hit rate as high as 97%, and even under these favorable conditions, the number of available tokens in the new plan is only approximately one-third of that in the old plan.68% similarUnverifiedInput tokens now account for more than 90% of total non-cached tokens, and the cost share of input context has risen from 47.5% to nearly 70%.67% similarUnverified在1.78百万token的长会话实测中，因为97%的token被缓存，发送token最多的配置反而最便宜67% similarUnverifiedThe user's calculation already accounts for a cache hit rate as high as 97% in their own usage scenario.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/926999API
curl https://kongchang.com/api/v1/knowledge/claims/926999MCP
get_claim(id=926999)