All models

DeepSeek V4 Flash 0731

deepseek/deepseek-v4-flash-0731

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....

Modalities
Text
In / Out price
$0.08 / $0.18 per 1M
Context
1.31M · 384K out
Released
Jul 31, 2026

Providers

Live list prices and uptime across 28 providers serving this model.

ProviderInput /MOutput /MCache read /MContextMax outputUptime 24h
OpenInferencefp8$0.03$0.07$0.0051.05M944K99.80%
Baidufp8$0.04$0.119$0.0011.05M131K99.93%
Relacefp4$0.04$0.08$0.0081.05M944K99.50%
Inceptronfp4$0.056$0.127$0.0041.05M944K99.84%
DeepInfrafp8$0.06$0.18$0.0151.05M384K99.70%
StreamLakefp8$0.062$0.185$0.0021.02M384K99.71%
Sail Researchfp4$0.078$0.36$0.0241.05M944K98.86%
DigitalOcean$0.08$0.252$0.0251.05M944K97.66%
Makora$0.09$0.195$0.021M384K99.35%
Wafer$0.10$0.25$0.051.05M944K99.89%
Rekafp4$0.11$0.66$0.007262K131K99.77%
Morphbf16$0.123$0.347$0.0311.05M944K99.99%
BaseTenfp8$0.13$0.26$0.0281.05M944K99.84%
BaseTenfp8$0.13$0.26$0.0281.05M384K99.82%
CoreWeavefp8$0.13$0.28$0.07262K236K99.77%
Together$0.14$0.28$0.031.05M944K99.76%
Parasailfp8$0.14$0.28$0.051.05M944K99.99%
Venice$0.175$0.35$0.0351M33K99.37%
Mancer 2fp8$0.175$0.501.05M944K99.32%
Fireworks$0.22$0.66$0.0071.05M944K97.36%
SiliconFlowfp8$0.22$0.66$0.0281.05M393K99.57%
GMICloudfp4$0.286$0.858$0.0091.05M944K99.73%
NextBitfp8$0.352$1.056$0.0121.05M944K99.91%
Alibaba$0.352$1.056$0.0351M393K99.99%
Novitafp8$0.409$1.228$0.0261.05M393K99.99%
Phala$0.44$1.32$0.0281.05M393K99.67%
AtlasCloudfp4$0.44$1.32$0.0281.05M393K99.28%
Cloudflare$0.44$1.32$0.0141.31M1.18M99.98%

No implicit prompt caching; cache reads require explicit cache control.

Capabilities

Tool calling
Reasoning
frequency_penalty
include_reasoning
logit_bias
logprobs
max_tokens
min_p
parallel_tool_calls
presence_penalty
reasoning
reasoning_effort
repetition_penalty
response_format
seed
stop
structured_outputs
temperature
tool_choice
top_a
top_k
top_logprobs
top_p

Pricing detail

Input
$0.08
Output
$0.18
Cache read
$0.016
Cache write

USD per 1M tokens.