DeepSeek
Efficient open and commercial models at competitive pricing.
DeepSeek develops highly efficient MoE and reasoning models that rival frontier closed models at a fraction of the cost, with open-weight options available.
Models
5
in catalog
Avg Input Price
$0.306
per 1M tokens
Avg Output Price
$1.174
per 1M tokens
Price Range
$0.14 – $0.55
input / 1M
DeepSeek Pricing History
Average input and output pricing across model portfolio
DeepSeek Release Timeline
Recent releases and capability milestones
Reasoning support added
capability
Functions support added
capability
Functions support added
capability
DeepSeek V3 released
release
DeepSeek R1 released
release
Functions support added
capability
DeepSeek V3 0324 released
release
DeepSeek V4 cataloged
release
Model Catalog
5 models currently tracked
| Model | Context | Input | Coding | Tier |
|---|---|---|---|---|
| DeepSeek V4 | 163.8K | $0.30 | 92 | fast |
| DeepSeek V3 0324 | 128K | $0.27 | 92 | standard |
| DeepSeek R1 | 64K | $0.55 | 94 | reasoning |
| DeepSeek V3 | 64K | $0.27 | 91 | standard |
| DeepSeek Coder V2 | 128K | $0.14 | 93 | fast |
Related Timeline Events
Recent activity from the AI release timeline
DeepSeek V4
Model Release — DeepSeek V4 launched by DeepSeek. Next-gen DeepSeek general model emphasizing co…
DeepSeek V4
API Release — API endpoint `deepseek-v4` available for DeepSeek V4.…
DeepSeek V4
Pricing Update — Initial pricing set at $0.30/M input and $1.20/M output tokens.…
DeepSeek V4
Function Calling — DeepSeek V4 supports function calling and structured tool use.…
DeepSeek V4
Reasoning Model — DeepSeek V4 introduces extended reasoning and chain-of-thought capabilities.…
DeepSeek V4
Benchmark Milestone — DeepSeek V4 achieves 92/100 coding benchmark score — top-tier performance.…
Compare with Competitors
- DeepSeek R1 vs GPT-5(DeepSeek vs OpenAI)
- DeepSeek R1 vs Claude 4 Opus(DeepSeek vs Anthropic)
- DeepSeek R1 vs Gemini 3 Pro(DeepSeek vs Google)
Related Calculators
- AI Cost Calculator
Estimate daily, monthly, and yearly API costs with provider and model selection.
- Context Window Utilization
Calculate how much of a model's context window you're consuming.
- Inference Latency Estimator
Project response times based on output tokens and throughput.
- Estimate DeepSeek API costs →
Frequently Asked Questions
- Why is DeepSeek pricing so competitive?
- DeepSeek uses mixture-of-experts architecture and efficient training to deliver frontier-class performance at lower API costs.
- What is DeepSeek R1?
- DeepSeek R1 is an open reasoning model with o1-class performance for math, coding, and STEM tasks.