DeepSeek V4 Flash 0731 — provider prices

29 providers offer this model. Cheapest input is $0.059, priciest is $0.440 — a 7.5x spread.

Providers
29
Cheapest
$0.059
Priciest
$0.440
Spread
7.5x

All quotes, cheapest first

ProviderInputOutput ContextQuantization
Wafer$0.059$0.3001049Knvfp4
DeepInfra$0.060$0.1801049KFP8 — good balance
Relace$0.070$0.1401049KFP4 — lowest precision, expect quality loss
Relace$0.070$0.1401049KFP4 — lowest precision, expect quality loss
Wafer$0.085$0.2301049Knvfp4
PrimeIntellect$0.090$0.2301049KFP4 — lowest precision, expect quality loss
Reka$0.100$0.280262Knot disclosed
DigitalOcean$0.119$0.2381049Knot disclosed
BaseTen$0.130$0.2601049KFP8 — good balance
BaseTen$0.130$0.2601049Kfp32
Venice$0.130$0.2601000Knot disclosed
CoreWeave$0.130$0.280262KFP8 — good balance
Inceptron$0.130$0.4951049KFP4 — lowest precision, expect quality loss
StreamLake$0.132$0.3961024KFP8 — good balance
Cohere$0.140$0.2801049Knot disclosed
Parasail$0.140$0.2801049KFP8 — good balance
Sail Research$0.150$0.3001049KFP4 — lowest precision, expect quality loss
Sail Research$0.150$0.3001049KFP4 — lowest precision, expect quality loss
Alibaba$0.176$0.5281000Knot disclosed
Alibaba$0.187$0.5611000Knot disclosed
Mancer 2$0.200$0.6001049KFP8 — good balance
SiliconFlow$0.220$0.6601049KFP8 — good balance
GMICloud$0.286$0.8581049KFP8 — good balance
Phala$0.308$0.9241049Knot disclosed
Novita$0.409$1.231049KFP8 — good balance
Chutes$0.440$1.321049KFP8 — good balance
Baidu$0.440$1.321049KFP8 — good balance
AtlasCloud$0.440$1.321049KFP4 — lowest precision, expect quality loss
Cloudflare$0.440$1.321049Knot disclosed

USD per million tokens. Verified 2026-11-04. Source: OpenRouter public API.

Cheapest is Wafer at $0.059 input. Check the quantization column before switching: undisclosed or FP4 quotes can produce noticeably worse output on reasoning and code tasks.

Related

All DeepSeek models · Price ranges · Cost calculator · Official pricing