DeepSeek’s official website is deepseek.com. Use chat.deepseek.com for the chat service. Developers can manage API access through the DeepSeek API platform and consult the official API documentation.
Which DeepSeek destination should you use?
- Main website:
deepseek.com - Chat:
chat.deepseek.com - API access and keys:
platform.deepseek.com - Developer documentation:
api-docs.deepseek.com
A third-party page mentioning DeepSeek does not establish that it is operated by DeepSeek. For official destinations, follow links from DeepSeek’s own website.
Choosing a DeepSeek model for production API use
Start with your workload, then check the official model list and model pricing page before choosing an API model. The API docs list deepseek-flash and deepseek-v4-pro. The model-list endpoint, GET https://api.deepseek.com/models, returns model identifiers and metadata that can include context window, output limit, supported input and output modalities, effort levels, and API capabilities. Metadata may be omitted when unavailable for a model.
For image-input tasks such as describing pictures, reading screenshot text, or analyzing charts, DeepSeek’s vision guide documents image input for deepseek-flash. The pricing page also marks tool calls and JSON output as supported for both listed models. Check the linked guides for implementation details and any feature-specific conditions, then test representative prompts for task quality and latency before deployment.
DeepSeek’s September 14, 2026 notice says deepseek-v4-pro requests would route to V4.1-Flash at Flash rates until V4.1-Pro launches. The current pricing table lists deepseek-v4-pro with model version DeepSeek-V4-Pro-0813, but the available documentation does not establish whether that routing notice has ended. Do not infer the backend from the model name alone; check the live documentation and validate the behavior you require before relying on it in production.
API pricing to consider
DeepSeek lists model prices in $ per 1 million tokens, with separate rows for cached input, uncached input, and output. The pricing table lists deepseek-flash uncached input at $0.15 off-peak and $0.30 at peak; output is $0.60 off-peak and $1.20 at peak. For deepseek-v4-pro, uncached input is $0.66 off-peak and $1.32 at peak; output is $1.98 off-peak and $3.96 at peak. Cached-input rates are also listed separately: Flash is $0.003 off-peak and $0.006 at peak; Pro is $0.022 off-peak and $0.044 at peak.
The pricing page says off-peak rates are half of peak rates. Peak hours are 01:00–04:00 and 06:00–10:00 UTC, Monday through Friday, excluding Chinese public holidays; all other hours, including weekends and Chinese public holidays, are off-peak. DeepSeek says prices may vary and recommends checking the page for the latest pricing. Use your expected input, output, and cache usage to estimate spend, and recheck the official API pricing page before budgeting or deployment.
For API setup, authentication, and request examples, use DeepSeek’s first API call guide.
Sources
DeepSeek API Documentation — Your First API Call
Lists Models | DeepSeek API Docs
Models & Pricing | DeepSeek API Docs
Tool Calls | DeepSeek API Docs