DeepSeek

DeepSeek V4 Flash

Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work

deepseek/deepseek-v4-flash

Context

1M

Output

384K

Released

2026-04-24

Knowledge cutoff

2025-05

Types

InputText
OutputText

Reasoning

Supported

Tool use

Supported

Structured output

Supported

Attachments

—

Open weights

Supported

Provider pricing

Kilo Gateway
InText
OutText
Price
$0.09 / $0.18
Context
1M
Status
Stable
OpenRouter
InText
OutText
Price
$0.09 / $0.18
Context
1M
Status
Stable
TokenGo
InText
OutText
Price
$0.10 / $0.20
Context
1M
Status
Stable
Vercel AI Gateway
InText
OutText
Price
$0.13 / $0.26
Context
1M
Status
Stable
Ambient
InText
OutText
Price
$0.14 / $0.28
Context
1M
Status
Stable
HPC-AI
InText
OutText
Price
$0.14 / $0.28
Context
1M
Status
Stable
NanoGPT
InText
OutText
Price
$0.14 / $0.28
Context
1M
Status
Stable
ZenMux
InText
OutText
Price
$0.14 / $0.28
Context
1M
Status
Stable
OrcaRouter
InText
OutText
Price
$0.15 / $0.30
Context
1M
Status
Stable
CommandCode
InText
OutText
Price
$0.15 / $0.60
Context
1M
Status
Stable
Eden AI
InTextImage
OutText
Price
$0.15 / $0.60
Context
1M
Status
Stable
Impossibl
InText
OutText
Price
$0.19 / $0.51
Context
1M
Status
Stable
CrossModel
InText
OutText
Price
$0.30 / $1.20
Context
1M
Status
Stable
Merge Gateway
InText
OutText
Price
$0.30 / $1.20
Context
1M
Status
Stable
AI/ML API
InText
OutText
Price
$0.39 / $1.56
Context
1M
Status
Stable
Ofox
InText
OutText
Price
$0.44 / $1.32
Context
1M
Status
Stable
Prime Inference
InText
OutText
Price
$0.44 / $1.32
Context
1M
Status
Stable
AnyAPI
InText
OutText
Price
—
Context
1M
Status
Stable

Benchmarks

  • SWE-Bench Verified

    79 · resolved

  • MMLU-Pro

    86.2 · EM

  • SimpleQA-Verified

    34.1 · pass@1

  • Chinese SimpleQA

    78.9 · pass@1

  • GPQA Diamond

    88.1 · pass@1

  • Humanity's Last Exam

    34.8 · pass@1

  • LiveCodeBench

    91.6 · pass@1

  • Codeforces

    3052 · rating

  • HMMT

    94.8 · pass@1

  • IMOAnswerBench

    88.4 · pass@1

  • MathArena Apex

    33 · pass@1

  • MathArena Apex Shortlist

    85.7 · pass@1

  • MRCR

    78.7 · MMR

  • CorpusQA

    60.5 · accuracy

  • Terminal-Bench

    56.9 · accuracy

  • SWE-Bench Pro

    52.6 · resolved

  • SWE-Bench Multilingual

    73.3 · resolved

  • BrowseComp

    73.2 · pass@1

  • Humanity's Last Exam

    45.1 · pass@1

  • MCP Atlas

    69 · pass@1

  • GDPval-AA

    1395 · Elo

  • Toolathlon

    47.8 · pass@1

From the same lab