DeepSeek: DeepSeek V3 API Pricing

by deepseek · 163K context window ·

$0.320
Input /1M tokens
$0.890
Output /1M tokens
163K
Context window

Explore DeepSeek V3 API pricing for your large language model projects. Understand the cost implications with transparent rates: $0.32000 per 1 million input tokens and $0.89000 per 1 million output tokens, totaling $1.2100 per million tokens. DeepSeek offers a substantial 163,840 token context window. Ideal for ML engineers evaluating LLM API costs, this data-driven pricing allows for informed decisions. For instance, a project processing 50 million input tokens and generating 10 million output tokens monthly would cost approximately $25. Explore if DeepSeek's pricing aligns with your project's budgetary needs.

Monthly Cost Examples

Assuming 50% input / 50% output token split

UsageMonthly cost
100K tokens/month$0.06
1M tokens/month$0.60
10M tokens/month$6.05
100M tokens/month$60.50

Compare with other models

DeepSeek: DeepSeek V3 vs OpenAI: GPT-4o →DeepSeek: DeepSeek V3 vs OpenAI: GPT-4o-mini →DeepSeek: DeepSeek V3 vs OpenAI: o1 →

Automate your model selection

StormRouter sends each request to the cheapest model that can handle it.
Only use DeepSeek: DeepSeek V3 when your quality requirements demand it.

Try StormRouter free →

Similar models

DeepSeek: R1 Distill Qwen 32B$0.580/1MAllenAI: Olmo 3.1 32B Think$0.650/1MDeepSeek: DeepSeek V3.2$0.650/1MAllenAI: Olmo 3 32B Think$0.650/1MQwen: Qwen3 VL 30B A3B Instruct$0.650/1MDeepSeek: DeepSeek V3.2 Exp$0.680/1M