免费
DeepSeek
Open-source AI Assistant from China
关于 DeepSeek
DeepSeek is a powerful open-source AI assistant developed in China, offering competitive performance with leading AI models. Known for its strong reasoning and coding capabilities, DeepSeek provides cost-effective AI solutions for developers and businesses. It supports multiple languages, offers API access, and excels at complex problem-solving and code generation tasks.
349460266 monthly visits
% popularity
定价
FreeVisit Trend
Key Metrics
2026-07-01 - 2026-07-31- Monthly visits
- 349460266
- Avg. visit duration
- 03:27
- Pages per visit
- 2.95
- Bounce rate
- 40.56%
Source Share
Direct visits: 44%Email: 3%Organic search: 34%Paid ads: 8%Referrals: 11%
Traffic Sources
2026-07-01 - 2026-07-31- Direct visits154,713,044
- Email10,462,662
- Organic search118,263,980
- Paid ads27,449,072
- Referrals38,571,508
DeepSeek 核心功能
DeepSeek-V4-Pro Flagship Model — Latest DeepSeek-V4-Pro-0813 with stronger Agent capabilities, Responses API support, and Codex integration — available on web, app, and API.
Thinking Mode (DeepThink) — Supports both non-thinking and thinking modes. Switch reasoning effort for complex problem-solving, coding, and multi-step agent tasks.
1M Context & 384K Max Output — Ultra-long context window up to 1 million tokens with maximum output of 384K tokens for large documents and long conversations.
OpenAI & Anthropic Compatible API — Drop-in compatible with OpenAI and Anthropic SDKs. Use base URL https://api.deepseek.com (or /anthropic) with minimal code changes.
Tool Calls, JSON Output & Caching — Built-in tool calling, structured JSON output, Responses API, chat prefix completion, and automatic context caching for cheaper repeated prefixes.
Free Web & App Chat — Chat for free on deepseek.com and the DeepSeek mobile app — experience V4 models without a subscription. API usage is billed separately by tokens.
DeepSeek 订阅计划
Free Chat
0$
- Free access on web and DeepSeek App
- DeepSeek-V4-Pro & V4-Flash experience
- Thinking mode (DeepThink) supported
- No monthly subscription required
V4-Flash API
from $0.22/M
- Model: deepseek-v4-flash (DeepSeek-V4-Flash-0731)
- Off-peak: $0.007 / $0.22 / $0.66 per 1M (cache hit / miss / output)
- Peak: $0.014 / $0.44 / $1.32 per 1M tokens
- 1M context, 384K max output, concurrency up to 2500
- JSON output, tool calls, Responses API
V4-Pro API
from $0.66/M
- Model: deepseek-v4-pro (DeepSeek-V4-Pro-0813)
- Off-peak: $0.022 / $0.66 / $1.98 per 1M (cache hit / miss / output)
- Peak: $0.044 / $1.32 / $3.96 per 1M tokens
- Stronger Agent capabilities; concurrency up to 500
- Peak hours: 01:00–04:00 & 06:00–10:00 UTC (off-peak is half price)
