DeepSeek V4 Flash

by DeepSeek

DeepSeek V4 Flash on DeepInfra with hybrid attention and reasoning modes

ReasoningToolsStructured output

Context window

1.0M

Max output

—

Input price

$0.14 /1M

Output price

$0.28 /1M

Benchmarks

Independent benchmark scores — composite indices for reasoning, coding, and math, plus individual eval scores where available.

Global rank#37 of 505 LLMs

TierStrong

Output speed72 tok/s

First token0.82s

Intelligence Index46.5

Coding Index38.7

GPQA Diamond

89%

Humanity's Last Exam

32%

SciCode

45%

IFBench

79%

Long-context reasoning

63%

Terminal-Bench Hard

36%

τ²-Bench Telecom

95%

See full leaderboard →Benchmarks via Artificial Analysis · View on AA

Available routes

The same DeepSeek V4 Flash runs on 2 different routes through the Opper gateway. Each route has its own data-handling posture.

Provider	Region	Zero data retention	Training	Logging	Contract	Input	Output
DeepInfra	US	—	No	Pending verification	PAYG	$0.14	$0.28
Novita	US	None	—	Not published; appendix available upon request.	PAYG	$0.14	$0.28

Human review: Abuse-flagged only

Other models from DeepSeek