DeepSeek

DeepSeek V4 Flash — Benchmark Scores, Pricing & Performance Analysis

OPEN SOURCEDeepSeek
Input Cost
$0.04/1M
Output Cost
$0.11/1M
Context Window
1.3M
Max Output
384K
Knowledge Cutoff
May 2025
Accepts
text, image

DeepSeek V4 Flash by DeepSeek demonstrates competitive pricing. View detailed benchmark data including scores across coding, math, reasoning, speed, and cost metrics.

Specifications last checked against the provider on . Benchmark scores carry their own source, linked beside each number.

Coding Benchmarks

Website Arena ELO
122044th of 108

Human preference ELO for building a working web page from a brief

Design Arena (via OpenRouter)
Design Arena ELO
118948th of 120

Human preference ELO for building things — websites, UI, games, charts, SVG

Design Arena (via OpenRouter)

Cost Benchmarks

Input Cost
$0.0480th of 413

Cost per 1M input tokens

OpenRouter API
Output Cost
$0.1193rd of 413

Cost per 1M output tokens

OpenRouter API
Cached Input Cost
$0.001st of 136

Cost per 1M cached input tokens — the price that actually applies to a long agent conversation

OpenRouter API

Context Benchmarks

Context Length
1.3M4th of 451

Maximum context window size

OpenRouter API

Available from 65 providers

The same model costs different amounts depending on who serves it. You can bring your own key for 6 of these — connect it here.

ProviderInput / 1MOutput / 1MCached inContextYour key
Alibaba Token Plan$0$0$01.0M
Alibaba Token Plan (China)$0$0$01.0M
InferX$0$01.0M
Kenari$0$01.0M
Pendra$0$01.0M
SCNet Token Plan$0$0$01.0M
SenseNova (China)$0$0$01.0M
Umans AI Coding Plan$0$0$01.0M
Volcengine Ark Coding Plan$0$0$01.0M
Merge Gateway$0.035$0.07$0.0071.0M
DevPass (LLM Gateway)$0.05$0.1$0.011.1M
UnoRouter$0.0625$0.1251.0M
DigitalOcean$0.0679$0.168$0.01681.0M
LLM Gateway$0.076$0.153$0.0141.0M
OpenRouter$0.088606$0.177212$0.0177211.0MSupported
Deep Infra$0.09$0.18$0.0181.0MSupported
TokenGo$0.098$0.196$0.0281.0M
Modelis$0.0983$0.19661.0M
Pioneer$0.1$0.2$0.01971.0M
GMI Cloud$0.112$0.224$0.0221.0M
routing.run$0.112$0.2241.0M
CrofAI$0.12$0.21$0.0031.0M
Vercel AI Gateway$0.13$0.26$0.0281.0MSupported
Abacus$0.14$0.28$0.031.0M
AIHubMix$0.14$0.28$0.0281.0M
Alibaba (China)$0.14$0.28$0.00281.0M
Ambient$0.14$0.28$0.0281.0M
Arcee$0.14$0.28$0.0281.0M
Auriko$0.14$0.28$0.00281.0M
ClinePass$0.14$0.28$0.00281.0M
EmpirioLabs AI$0.14$0.28$0.141.0M
HPC-AI$0.14$0.28$0.0281.0M
Hugging Face$0.14$0.281.0MSupported
Jalapeno Cloud$0.14$0.281.0M
Kilo Gateway$0.14$0.28$0.0281.0M
NanoGPT$0.14$0.28$0.00281.0M
Neuralwatt$0.14$0.28$0.0281.0M
NovitaAI$0.14$0.28$0.0281.0M
Nvidia$0.14$0.28$0.00281.0M
OpenCode Zen$0.14$0.28$0.0281.0M
SiliconFlow$0.14$0.28$0.0281.0M
SiliconFlow (China)$0.14$0.28$0.0031.0M
Umans AI$0.14$0.28$0.0281.0M
Weights & Biases$0.14$0.28$0.071.0M
ZenMux$0.14$0.28$0.00281.0M
EBCloud$0.143$0.28571.0M
OrcaRouter$0.147$0.295$0.021.0M
ai&$0.15$0.251.0M
DeepSeek$0.15$0.6$0.0031.0MSupported
OpenCode Go$0.15$0.6$0.0031.0M
TensorX$0.15$0.3$0.03751.0M
Vivgrid$0.15$0.3$0.031.0M
above.dev$0.165$0.66$0.00331.0M
Azure$0.19$0.511.0MSupported
Impossibl$0.19$0.51$0.0281.0M
Charm Hyper$0.2$0.4$0.041.0M
Vancine$0.22$0.66$0.0071.0M
CrossModel$0.27$1.08$0.00541.0M
Requesty$0.28$0.56$0.071.0M
Vultr$0.3$11.0M
Eden AI$0.44$1.32$0.0141.0M
Ofox$0.44$1.32$0.0141.0M
AgentRouter1.0M
AnyAPI1.0M
Ollama Cloud1.0M

DeepSeek V4 Flash — Benchmark Scores Overview

Scores normalized to percentage scale for visual comparison. ELO scores mapped to 0-100 range (1100-1500).

DeepSeek V4 Flash — Frequently Asked Questions

How much does DeepSeek V4 Flash cost?

DeepSeek V4 Flash costs $0.04 per 1M input tokens and $0.11 per 1M output tokens. This makes it one of the more affordable models.

What is the context window of DeepSeek V4 Flash?

DeepSeek V4 Flash has a context window of 1.3M tokens. This determines how much text, conversation history, and code the model can process in a single request.

Who created DeepSeek V4 Flash?

DeepSeek V4 Flash was created by DeepSeek. It is classified as a open source model in our catalogue.

Is DeepSeek V4 Flash open source?

Yes, DeepSeek V4 Flash is an open-source model. The model weights are publicly available for download and self-hosting.