Cost levers - verified
How to cut your DeepSeek API bill
Prompt caching: automatic. Cache lifetime: hours to days, cleared once unused. A cache hit typically saves 97.3% on input across 2 current models. Batch: no batch API documented. Service tiers: none documented.
Read from api-docs.deepseek.com. Compare every provider's cost levers
- Caching
- Automatic
- Cache hit saves
- 97.3%
- Batch
- None
- Service tiers
- 0
Prompt caching
- Mode
Automatic: repeated prefixes are cached with no code change
Context Caching on Disk, enabled by default; hits require a full match of a persisted cache prefix unit (request boundaries, detected common prefixes, fixed token intervals).
- Cache lifetime
Automatically cleared once unused, usually within a few hours to a few days
- Minimum prompt
- Not stated
- Cache write cost
- Not stated
- Cache read cost
Per-model cache-hit price (also halved off-peak)
- Storage cost
- Not stated
- How to enable
On by default; no code changes. Check usage.prompt_cache_hit_tokens / prompt_cache_miss_tokens.
- From the docs
“The DeepSeek API Context Caching on Disk Technology is enabled by default for all users, allowing them to benefit without needing to modify their code.”
api-docs.deepseek.com “Cache construction takes seconds. Once the cache is no longer in use, it will be automatically cleared, usually within a few hours to a few days.”
api-docs.deepseek.com “The cache system works on a "best-effort" basis and does not guarantee a 100% cache hit rate.”
api-docs.deepseek.com
Batch API
No batch API documented
No batch API on the pricing page or caching guide.
Service tiers
Flex trades speed or availability for a lower price; Priority and Fast cost more for lower latency or reserved capacity.
Service tiers on the pages checked: Not stated
Other discounts and surcharges
Off-peak pricing (50% of peak)
Peak = 01:00-04:00 and 06:00-10:00 UTC Monday to Friday, excluding Chinese public holidays; all other hours off-peak (including weekends and Chinese public holidays in full)
“Off-peak rates are half of the peak rates. Peak hours are 01:00 - 04:00 and 06:00 - 10:00 UTC, Monday through Friday, excluding Chinese public holidays. All other hours are off-peak, including weekends and Chinese public holidays in full.”
api-docs.deepseek.com
DeepSeek per-model prices
USD per 1M tokens from the provider's own pricing page, newest release first. A dash means the model has no such rate. Muted rows are scheduled for shutdown or deprecated.
| Model | Input | Cached input | Batch input | Output | Batch output |
|---|---|---|---|---|---|
| DeepSeek V4.1 Flash | $0.30 | $0.00698% off | - | $1.20 | - |
| DeepSeek V4 Pro | $1.32 | $0.04496.7% off | - | $3.96 | - |
Research notes
facts.json records DeepSeek peak rates. Legacy deepseek-v4-flash names now route to DeepSeek-V4.1-Flash-1017 billed at Flash price. DeepSeek's own non-model API line items on the pricing page excluded.
FAQ
Does DeepSeek cache prompts automatically?
Automatic: repeated prefixes are cached with no code change. Context Caching on Disk, enabled by default; hits require a full match of a persisted cache prefix unit (request boundaries, detected common prefixes, fixed token intervals). How to enable: On by default; no code changes. Check usage.prompt_cache_hit_tokens / prompt_cache_miss_tokens.
How much does a cache hit cost on DeepSeek?
Across 2 current models with a cache read price, a hit saves 97.3% (96.7% to 98%) off the input price. The docs say: Per-model cache-hit price (also halved off-peak).
Does DeepSeek have a batch API discount?
No batch API is documented. No batch API on the pricing page or caching guide.
Sources
Last verified from DeepSeek's own pages below. Quotes are verbatim; each was matched against the live page.
Newsletter
Know when a provider changes its prices
One email per week. New model releases, pricing changes and what they mean for what you build.