DeepSeek: DeepSeek V4 Pro
deepseek/deepseek-v4-pro
30th smartest of 182ReasoningToolsJSON
Intelligence
44.3
30th of 182
Design Elo
1317
3D
Speed
—
tokens/sec
Latency
—
first token
Input price
$0.435
164th cheapest
Context
1.0M
384K max out
How it compares
Smarter than84%
of all ranked models
Cheaper than51%
of all ranked models
Overview
DeepSeek V4 Pro is a large-scale Mixture-of-Experts model from DeepSeek with 1.6T total parameters and 49B activated parameters, supporting a 1M-token context window. It is designed for advanced reasoning, coding,...
Benchmarks
independent · Artificial Analysis & Design ArenaArtificial Analysis94th percentile
Intelligence Index
44.3
Coding Index
59.4
GPQA Diamond
89%
Humanity's Last Exam
36%
SciCode
50%
Tau²-Bench (agentic)
96%
Design Arena · Elo3,271 tournaments
3D
1317
Website
1260
Data Viz
1226
svg
1182
Providers & pricing (18)
| Provider | In $/M | Out $/M | Context | Uptime |
|---|---|---|---|---|
| DeepSeekcache | $0.435 | $0.870 | 1.0M | 100% |
| StreamLakefp8 | $0.670 | $1.34 | 1.0M | 99.3% |
| GMICloudfp8 | $0.679 | $1.36 | 1.0M | 99.7% |
| Ionstreamfp4 | $1.13 | $2.26 | 1.0M | 98.1% |
| Novitafp8 | $1.17 | $2.34 | 1.0M | 99.9% |
| DeepInfrafp4 | $1.30 | $2.60 | 1.0M | 99.7% |
| DigitalOcean | $1.39 | $2.78 | 262K | 98.8% |
| Alibabafp8 | $1.42 | $2.83 | 1M | 100% |
| SiliconFlowfp8 | $1.50 | $3.13 | 1.0M | 99.8% |
| Venice | $1.65 | $3.30 | 1M | 99.5% |
| AtlasCloudfp4 | $1.68 | $3.38 | 1.0M | 99.8% |
| Baidufp8 | $1.69 | $3.38 | 1.0M | 99.9% |
| BaseTenfp4 | $1.74 | $3.48 | 262K | 53.4% |
| Parasailfp8 | $1.74 | $3.48 | 1.0M | 99.9% |
| Cloudflare | $1.74 | $3.48 | 393K | 100% |
| Together | $1.74 | $3.48 | 512K | 94.7% |
| CoreWeavefp8 | $1.74 | $3.48 | 1.0M | 99.3% |
| Fireworks | $1.74 | $3.48 | 1.0M | — |
Specifications
Context window1.0M
Max output384K
Knowledge cutoff—
Input modalitiestext
Output modalitiestext
Prompt cachingSupported
Cache read price$0.0036/M
ModeratedNo
Open weightsdeepseek-ai/DeepSeek-V4-Pro ↗
DeepSeek V4 Pro FAQ
How much does DeepSeek V4 Pro cost?
DeepSeek V4 Pro costs $0.435 per million input tokens and $0.870 per million output tokens via OpenRouter, making it 164th cheapest of 332 paid models.
How smart is DeepSeek V4 Pro?
DeepSeek V4 Pro scores 44.3 on the Artificial Analysis Intelligence Index, ranking 30th of 182 benchmarked models, with a GPQA Diamond score of 89%.
What is DeepSeek V4 Pro's context window?
DeepSeek V4 Pro supports a 1.0M-token context window and can output up to 384K tokens. It accepts text input.
Similar models
All DeepSeek V4 Pro alternatives →Compare head-to-head
DeepSeek: DeepSeek V4 Pro vs OpenAI: GPT-5.3-CodexDeepSeek: DeepSeek V4 Pro vs MoonshotAI: Kimi K2.6DeepSeek: DeepSeek V4 Pro vs MiniMax: MiniMax M3DeepSeek: DeepSeek V4 Pro vs MiniMax: MiniMax M3 (batch)DeepSeek: DeepSeek V4 Pro vs Qwen: Qwen3.7 MaxDeepSeek: DeepSeek V4 Pro vs Xiaomi: MiMo-V2.5-Pro