DeepSeek V4 Flash — API Pricing & Benchmarks

DeepSeek V4 Flash is DeepSeek's mid reasoning model, released in 2026-07. The API costs $0.44 per 1M input tokens and $1.32 per 1M output tokens ($0.66 blended at a 3:1 ratio), with cached input at $0.01. It supports a 1M-token context window. On the independently measured Artificial Analysis Intelligence Index it scores 51.8, ranking #28 of 68 models tracked on this site, with a median output speed of 136 tokens/sec. Pricing was changed on 2026-08-16 (previously $0.14/$0.28 per 1M tokens).

Explore provider routes and additional benchmark sources for DeepSeek V4 Flash

ProviderDeepSeek
Input price / 1M tokens$0.44
Output price / 1M tokens$1.32
Cached input / 1M tokens$0.01
Blended price (3:1)$0.66
Context window1M
Max output tokens384K
Knowledge cutoff
Licenseopen weights
Input modalitiesT
Output speed (tok/s)136
Released2026-07
Artificial Analysis Intelligence Index51.8
LMArena Elo1432
GPQA Diamond90.8%
Humanity's Last Exam38.6%

Where DeepSeek V4 Flash fits

On a blended 3:1 basis DeepSeek V4 Flash costs $0.66 per 1M tokens, cheaper than 75% of the 80 priced models in this index. It sits in the mid-tier bracket where most production chat, RAG and summarization workloads run: meaningfully cheaper than flagships while keeping most of their capability. The weights are open, so it can be self-hosted for data-residency or cost reasons instead of consumed via API. Cached input is priced at $0.01 (97% below standard input), which rewards prompt structures with long stable prefixes — see our caching guide.

Price history

PeriodInput $/1MOutput $/1MCached $/1M
2026-07-18 → 2026-08-16$0.14$0.28$0.0028
2026-08-16 → current$0.44$1.32$0.01

List-price changes recorded by this site. The full dataset is public: price-history.json (CC BY 4.0).

Alternatives at a similar price

About DeepSeek

DeepSeek publishes open-weights models at disruptive prices. Even after its August 2026 switch to peak/off-peak billing, the V4 line costs a fraction of Western equivalents, and off-peak hours (everything outside 01:00–04:00 and 06:00–10:00 UTC on weekdays) halve the bill again.

More DeepSeek models

← Full comparison table (60+ models, live filters)