Unverified50% confidenceFactExact time
Fixed-quota plans (such as Volcano Coding Plan, Volcano Agent Plan and Baidu's Token-based plan) give a fixed quota per call, per AFP or per Token, with the same deduction whether or not the cache is hit, so caching saves nothing; three-tier coefficient plans (such as Alibaba, Tencent, Baidu's points-based plan, Zhipu, Xiaomi and MiniMax) deduct at separate coefficients for uncached input, cache hits and output, with cache hits deducted less, so the more caching the more you save.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
11/6/2026
First Seen
Valid until: 2/4/2027
Sources
Related Entities
Related Claims
UnverifiedCheap renewal services for Cursor on the market fall into roughly three categories: shared accounts and short-term Pro access, time-limited subscription packages, and cracked or modified client tools.69% similarVerifiedCached input tokens are a discount mechanism offered by providers such as OpenAI for large repeated prefixes (such as system prompts, long documents, and few-shot examples); input tokens that hit the cache are billed at a lower rate, usually only about one-tenth of the standard input price.69% similarUnverifiedSome API providers cache repeated context to reduce costs, but the caching strategy is opaque; the meter may still display usage by the full token count, or billing rules may differ depending on whether the cache hits.67% similarUnverifiedDifferent AI providers differ in cache granularity, hit conditions and billing methods; for example, some price cache-write tokens and cache-read tokens separately, and some set a minimum cacheable token length.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1142489API
curl https://kongchang.com/api/v1/knowledge/claims/1142489MCP
get_claim(id=1142489)