← Preços de API de LLM

DeepSeek V4 Flash 0731

DeepSeek

DeepSeek V4 Flash 0731 is a sparse mixture-of-experts model from DeepSeek, with 13B active parameters out of 284B total. This re-post-trained revision is suited for coding, reasoning, and agent workflows....

Ent.
$0.06
Saí.
$0.11
Misto
$0.07
Onde comprar este modelo
26 rotas de acesso
RotaEnt.Saí.Mistovs
OpenInference mais baratofp8$0.04$0.10$0.06-
Relace fp4$0.06$0.11$0.07+17%
StreamLake fp8$0.06$0.17$0.09+50%
DeepInfra fp8$0.06$0.18$0.09+50%
Inceptron fp4$0.06$0.17$0.09+50%
Makora $0.09$0.20$0.12+100%
Wafer $0.10$0.25$0.14+133%
Sail Research fp4$0.07$0.34$0.14+133%
DigitalOcean $0.12$0.24$0.15+150%
BaseTen fp8$0.13$0.26$0.16+167%
CoreWeave fp8$0.13$0.28$0.17+183%
Together $0.14$0.28$0.18+200%
Morph bf16$0.12$0.35$0.18+200%
Venice $0.17$0.35$0.22+267%
Reka fp4$0.11$0.66$0.25+317%
Mancer 2 fp8$0.20$0.60$0.30+400%
Fireworks $0.22$0.66$0.33+450%
SiliconFlow fp8$0.22$0.66$0.33+450%
GMICloud fp8$0.29$0.86$0.43+617%
NextBit fp8$0.35$1.06$0.53+783%
Alibaba $0.35$1.06$0.53+783%
Novita fp8$0.41$1.23$0.61+917%
Phala $0.44$1.32$0.66+1000%
Baidu fp8$0.44$1.32$0.66+1000%
AtlasCloud fp4$0.44$1.32$0.66+1000%
Cloudflare $0.44$1.32$0.66+1000%

API oficial · plataforma de nuvem · host de inferência · USD por 1M de tokens. Mesmo modelo, provedores diferentes — preço e quantização variam.

Ctx
1.31072M
max output
944K
cache read /1M
$0.01
Família
deepseek
Modalidades
text
Recursos
json · reasoning · tools

Preços ao vivo, atualizados 2026-09-15.

Conceitos nesta página