Ctrl K
Back to models

DeepSeek

DeepSeek-V4-Flash

  • Released
  • Open weights
  • MIT
Compare
Intelligence Index

Max setting

24.2
Measured · Artificial Analysis v4.3.2
API price from
$0.06 in / $0.14 out / 1M tokOpenRouter · OpenInference · FP8
Context window
1.05M tok

Performance profile

ReasoningArchived source
89.4%
GPQA Diamond
CodingArchived source
45.3%
SciCode
Cost efficiencyArchived source
204.9pts / $
Intelligence ÷ eval cost

Capabilities

Context window

Modalities

Reasoning modes

Integration

Reasoning trade-offs

Source (opens in a new tab)
Intelligence Index v4.3.2
DeepSeek-V4-Flash
Cost per task (USD) · log scale
DeepSeek-V4-FlashThink Max
Intelligence Index
24.2
Cost / task
$0.1179

Pricing & access

API route
OpenRouter · OpenInference · FP8API
Tier & contextStandard

USD per 1M tokens

Input
$0.06
Output
$0.14
Cached input
$0.014

Run locally

Original download
159.6 GB
App memory
Unverified
Documented software
vLLMSetup guide (opens in a new tab)
Find GGUF downloads (opens in a new tab)Original weights (opens in a new tab)

Benchmarks

Full leaderboardSource (opens in a new tab)

Humanity’s Last Exam

Frontier reasoning

DeepSeek-V4-Flash · MaxArchived

  1. 1Claude Opus 5.5Max61.4%
  2. 2Claude Fable 5.1Max59.1%
  3. 3Gemini 4 ArgonHigh57.1%
  4. 4Claude Sonnet 5.5Max55.0%
  5. 5GPT-6 AstraMax54.7%
  6. 6Muse Wonder FullMax53.6%
  7. 7GPT-6.1 SolMax52.9%
  8. 8Muse WonderMax50.1%
  9. 9MiMo-V2.6-ProReasoning49.4%
  10. 10Muse Spark 1.3Max48.7%

Peers use their best measured result on this benchmark.