DeepSeek API Pricing
DeepSeek models from DeepSeek. Chinese AI company developing advanced language models. Requesty routes to 4 DeepSeek models starting at $0.14 per 1M input tokens with context windows up to 1M tokens. One API key, OpenAI-compatible SDK, no markup.
Flagship model
deepseek-v4-proIntelligence Index
51.5
Coding Index
47.5
GPQA Diamond
88.8%
Terminal-Bench Hard
46.2%
All DeepSeek models
| Model | Context | Max Output | Input/1M | Output/1M | Capabilities | Coding |
|---|---|---|---|---|---|---|
deepseek-v4-pro | 1M | 384K | $0.43 | $0.87 | 🔧⚡ | 48 |
deepseek-v4-flash | 1M | 384K | $0.14 | $0.28 | 🔧⚡ | 39 |
deepseek-reasoner | 1M | 384K | $0.14 | $0.28 | 🔧⚡ | N/A |
deepseek-chat | 1M | 384K | $0.14 | $0.28 | 🔧⚡ | N/A |
About DeepSeek on Requesty
How many DeepSeek models are available through Requesty?
Requesty routes to 4 DeepSeek models including regional variants, with pricing synced in real time to the upstream provider.
What is the cheapest DeepSeek model?
The cheapest DeepSeek model starts at $0.14 per million input tokens. See the pricing column in the table below for full per-model rates.
Does Requesty add markup on DeepSeek pricing?
No. Requesty passes through exactly what DeepSeek charges. You pay the same per-token rates as going direct, plus you get smart routing, caching, analytics, and one unified API for 400+ models.
Is my data used to train DeepSeek models?
DeepSeek's training policy varies by product and tier. See their privacy policy for specifics, and contact Requesty for enterprise-grade data controls.
Where are DeepSeek models hosted?
DeepSeek models are hosted in 🇨🇳 China. Some models are available in additional regions through AWS Bedrock, Azure, or Google Vertex AI: filter by region on the DeepSeek rows in the models explorer.