All Models
DeepSeek v4 Pro
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
Available Providers (47)
| Provider | Model ID | Input Cost | Output Cost | Context | Max Output | Docs |
|---|---|---|---|---|---|---|
 UnoRouter | deepseek-v4-pro:free | $0/MTok | $0/MTok | 1M | 384K | |
 Kenari | deepseek-v4-pro | $0/MTok | $0/MTok | 1M | 384K | |
 Alibaba Token Plan | deepseek-v4-pro | $0/MTok | $0/MTok | 1M | 384K | |
 Alibaba Token Plan (China) | deepseek-v4-pro | $0/MTok | $0/MTok | 1M | 384K | |
 routing.run | deepseek-v4-pro | $0.35/MTok | $0.70/MTok | 1M | 64K | |
 CrofAI | deepseek-v4-pro | $0.35/MTok | $0.80/MTok | 1M | 131.1K | |
 EBCloud | DeepSeek-V4-Pro | $0.43/MTok | $0.86/MTok | 1M | 384K | |
 Pioneer | deepseek-ai/DeepSeek-V4-Pro | $0.43/MTok | $0.87/MTok | 1M | 131.1K | |
 LLM Gateway | deepseek-v4-pro | $0.43/MTok | $0.87/MTok | 1.1M | 384K | |
 Hugging Face | deepseek-ai/DeepSeek-V4-Pro | $0.43/MTok | $0.87/MTok | 1.0M | 393.2K | |
 Nvidia | deepseek-ai/deepseek-v4-pro | $0.43/MTok | $0.87/MTok | 1.0M | 393.2K | |
 Merge Gateway | deepseek/deepseek-v4-pro | $0.43/MTok | $0.87/MTok | 1M | 384K | |
 Auriko | deepseek-v4-pro | $0.43/MTok | $0.87/MTok | 1M | 384K | |
 Vivgrid | deepseek-v4-pro | $0.43/MTok | $0.87/MTok | 1M | 384K | |
 Kilo Gateway | deepseek/deepseek-v4-pro | $0.43/MTok | $0.87/MTok | 1.0M | 384K | |
 OpenRouter | deepseek/deepseek-v4-pro | $0.43/MTok | $0.87/MTok | 1.0M | 384K | |
 Vercel AI Gateway | deepseek/deepseek-v4-pro | $0.43/MTok | $0.87/MTok | 1M | 384K | |
 DeepSeek | deepseek-v4-pro | $0.43/MTok | $0.87/MTok | 1M | 384K | |
 Alibaba (China) | deepseek-v4-pro | $0.43/MTok | $0.87/MTok | 1M | 384K | |
 ZenMux | deepseek/deepseek-v4-pro | $0.43/MTok | $0.87/MTok | 1M | 384K | |
 CrossModel | deepseek/deepseek-v4-pro | $0.47/MTok | $0.94/MTok | 1M | 65K | |
 OrcaRouter | deepseek/deepseek-v4-pro | $0.56/MTok | $1.12/MTok | 1M | 384K | |
| UnoRouter | deepseek-v4-pro | $0.90/MTok | $1.80/MTok | 1M | 384K | |
 NanoGPT | deepseek/deepseek-v4-pro | $1.10/MTok | $2.20/MTok | 1.0M | 384K | |
 Deep Infra | deepseek-ai/DeepSeek-V4-Pro | $1.30/MTok | $2.60/MTok | 1.0M | 16.4K | |
 GMI Cloud | deepseek-ai/DeepSeek-V4-Pro | $1.39/MTok | $2.78/MTok | 1.0M | 384K | |
 Cortecs | deepseek-v4-pro | $1.55/MTok | $3.11/MTok | 1.0M | 384K | |
 NovitaAI | deepseek/deepseek-v4-pro | $1.60/MTok | $3.20/MTok | 1.0M | 393.2K | |
 EmpirioLabs AI | deepseek-v4-pro | $1.65/MTok | $3.30/MTok | 1M | 393.2K | |
 Venice AI | deepseek-v4-pro | $1.65/MTok | $3.30/MTok | 1M | 32.8K | |
 FrogBot | deepseek-v4-pro | $1.74/MTok | $3.48/MTok | 128K | 8.2K | |
 Abacus | deepseek-ai/DeepSeek-V4-Pro | $1.74/MTok | $3.48/MTok | 1M | 32.8K | |
 Fireworks AI | accounts/fireworks/models/deepseek-v4-pro | $1.74/MTok | $3.48/MTok | 1M | 384K | |
 Weights & Biases | deepseek-ai/DeepSeek-V4-Pro | $1.74/MTok | $3.48/MTok | 1.0M | 1.0M | |
 FastRouter | deepseek/deepseek-v4-pro | $1.74/MTok | $3.48/MTok | 1M | 384K | |
 Azure | deepseek-v4-pro | $1.74/MTok | $3.48/MTok | 1M | 384K | |
 OpenCode Zen | deepseek-v4-pro | $1.74/MTok | $3.84/MTok | 1M | 384K | |
 SiliconFlow | deepseek-ai/DeepSeek-V4-Pro | $1.74/MTok | $3.48/MTok | 1M | 384K | |
 HPC-AI | deepseek/deepseek-v4-pro | $1.74/MTok | $3.48/MTok | 1.0M | 128K | |
 OpenCode Go | deepseek-v4-pro | $1.74/MTok | $3.48/MTok | 1M | 384K | |
 DigitalOcean | deepseek-v4-pro | $1.74/MTok | $3.48/MTok | 1.0M | 393.2K | |
 Baseten | deepseek-ai/DeepSeek-V4-Pro | $1.74/MTok | $3.48/MTok | 262.1K | 262.1K | |
 Together AI | deepseek-ai/DeepSeek-V4-Pro | $1.74/MTok | $3.48/MTok | 512K | 384K | |
 Nebius Token Factory | deepseek-ai/DeepSeek-V4-Pro | $1.75/MTok | $3.50/MTok | 1M | 384K | |
 AnyAPI | deepseek/deepseek-v4-pro | —/MTok | —/MTok | 1M | 384K | |
 Model Oracle AI | deepseek-v4-pro | —/MTok | —/MTok | 1M | 384K | |
 Ollama Cloud | deepseek-v4-pro | —/MTok | —/MTok | 1.0M | 1.0M |
Capabilities
Reasoning
Tool Calling
Attachments
Open Weights
Structured Output