DeepSeek: DeepSeek V4 Flash
deepseek/deepseek-v4-flash
40th smartest of 182Cheaper than 78% of paidReasoningToolsJSON
Intelligence
40.3
40th of 182
Design Elo
1246
3D
Speed
—
tokens/sec
Latency
—
first token
Input price
$0.140
72nd cheapest
Context
1.0M
393K max out
How it compares
Smarter than78%
of all ranked models
Cheaper than78%
of all ranked models
Overview
DeepSeek V4 Flash is an efficiency-optimized Mixture-of-Experts model from DeepSeek with 284B total parameters and 13B activated parameters, supporting a 1M-token context window. It is designed for fast inference and...
Benchmarks
independent · Artificial Analysis & Design ArenaArtificial Analysis90th percentile
Intelligence Index
40.3
Coding Index
56.2
GPQA Diamond
89%
Humanity's Last Exam
32%
SciCode
45%
Tau²-Bench (agentic)
95%
Design Arena · Elo7,908 tournaments
3D
1246
Website
1232
svg
1200
Data Viz
1156
Providers & pricing (21)
| Provider | In $/M | Out $/M | Context | Uptime |
|---|---|---|---|---|
| DeepInfrafp4 | $0.090 | $0.180 | 1.0M | 99.2% |
| StreamLakefp8 | $0.092 | $0.185 | 1.0M | 99.7% |
| GMICloudfp8 | $0.094 | $0.188 | 1.0M | 99.6% |
| AkashMLfp8 | $0.098 | $0.196 | 1.0M | 87.1% |
| DigitalOcean | $0.112 | $0.224 | 262K | 99.1% |
| SiliconFlowfp8 | $0.130 | $0.280 | 1.0M | 99.2% |
| Alibabafp8 | $0.134 | $0.268 | 1M | 99.9% |
| Venice | $0.138 | $0.275 | 1M | 99.5% |
| Morph | $0.139 | $0.278 | 1.0M | 99.6% |
| Io Netfp8 | $0.139 | $0.278 | 33K | 99.9% |
| Ionstreamfp4 | $0.140 | $0.280 | 1.0M | 91.5% |
| Parasailfp8 | $0.140 | $0.280 | 1.0M | 99.3% |
| Fireworks | $0.140 | $0.280 | 1.0M | 98.2% |
| Novitafp8 | $0.140 | $0.280 | 1.0M | 99.9% |
| Ambientfp4 | $0.140 | $0.280 | 1.0M | 83.6% |
| Cloudflare | $0.140 | $0.280 | 384K | 100% |
| AtlasCloudfp4 | $0.140 | $0.280 | 1.0M | 99.8% |
| DeepSeekcache | $0.140 | $0.280 | 1.0M | 100% |
| Baidufp8 | $0.140 | $0.280 | 1.0M | 99.6% |
| CoreWeavefp8 | $0.140 | $0.280 | 1.0M | 99.6% |
| Mancer 2fp4 | $0.200 | $0.500 | 1.0M | 98.3% |
Specifications
Context window1.0M
Max output393K
Knowledge cutoff—
Input modalitiestext
Output modalitiestext
Prompt cachingSupported
Cache read price$0.028/M
ModeratedNo
Open weightsdeepseek-ai/DeepSeek-V4-Flash ↗
DeepSeek V4 Flash FAQ
How much does DeepSeek V4 Flash cost?
DeepSeek V4 Flash costs $0.140 per million input tokens and $0.280 per million output tokens via OpenRouter, making it 72nd cheapest of 332 paid models.
How smart is DeepSeek V4 Flash?
DeepSeek V4 Flash scores 40.3 on the Artificial Analysis Intelligence Index, ranking 40th of 182 benchmarked models, with a GPQA Diamond score of 89%.
What is DeepSeek V4 Flash's context window?
DeepSeek V4 Flash supports a 1.0M-token context window and can output up to 393K tokens. It accepts text input.
Similar models
All DeepSeek V4 Flash alternatives →Compare head-to-head
DeepSeek: DeepSeek V4 Flash vs Z.ai: GLM 5.1DeepSeek: DeepSeek V4 Flash vs OpenAI: GPT-5.2-CodexDeepSeek: DeepSeek V4 Flash vs Qwen: Qwen3.6 Max PreviewDeepSeek: DeepSeek V4 Flash vs OpenAI: GPT-5.4 MiniDeepSeek: DeepSeek V4 Flash vs OpenAI: GPT-5.4 Mini (batch)DeepSeek: DeepSeek V4 Flash vs Thinking Machines: Inkling