API Launches Context Caching on Disk Technology
Review official pricing for cache-hit rates after disk context caching launched.
DeepSeek's change log says: The DeepSeek API has innovatively adopted hard disk caching, reducing prices by another order of magnitude.
Who it affects: API users billed per token with cacheable prompts.
What it means: Review official pricing for cache-hit rates after disk context caching launched.
Official sources
Related pages: /api-pricing · /guides/off-peak-pricing · /news
Context for developers
Disk context caching lets DeepSeek serve repeated prompt prefixes from cheaper storage tiers, which shows up as separate cache-hit input pricing on the Models & Pricing page. The 2024-08-02 announcement describes another order-of-magnitude reduction versus uncached input — read the live table for today's exact dollars per million tokens.
Client SDKs must enable caching features documented in the API guides (prefix hashing, cache breakpoints). Billing still depends on whether a token hits cache; monitor usage dashboards after rollout. Pair with error codes if you see unexpected 402/429 responses when cache traffic spikes.
Independent resource, not affiliated with DeepSeek.