AI Compare AI Compare

Check prices, features and limits.
Pricing market
Automatic
Language
EN

DeepSeek

AI assistant from the Chinese vendor DeepSeek: chat via web and app plus a token-billed API for its own language models.

Free accessFree plan
Paid fromAPI deepseek-flashInput cache hit: $0.003 (off-peak) / $0.006 (peak) per 1M tokens

Category: Productivity · Assistants & workspaces

Features

Prices and plans

TierPricePeriodWhat you get
DeepSeek Web / App (Chat)contact salesFreeHome page verbatim: 'Free access to DeepSeek. Experience the intelligent model.'
Access via web chat (chat.deepseek.com) and the DeepSeek app
Per the home page DeepSeek-V4-Pro is 'fully available across the web, mobile app, and API'
The vendor publishes no paid subscription tier for the chat product
API deepseek-flashcontact sales1 million Entry tokens (cache miss, off time / peak time)Input cache hit: $0.003 (off-peak) / $0.006 (peak) per 1M tokens
Input cache miss: $0.15 (off-peak) / $0.30 (peak) per 1M tokens
Output: $0.60 (off-peak) / $1.20 (peak) per 1M tokens
Context length 1M, max output 384K, model version DeepSeek-V4.1-Flash
Concurrency limit 2500
Vendor note verbatim: 'Off-peak rates are half of the peak rates. Peak hours are 01:00 - 04:00 and 06:00 - 10:00 UTC, Monday through Friday'
Supersedes deepseek-v4-flash and deepseek-v4-flash-vision-exp; now also accepts image input (images are billed as input tokens)
API deepseek-v4-procontact sales1 million Entry tokens (cache miss, off time / peak time)Input cache hit: $0.022 (off-peak) / $0.044 (peak) per 1M tokens
Input cache miss: $0.66 (off-peak) / $1.32 (peak) per 1M tokens
Output: $1.98 (off-peak) / $3.96 (peak) per 1M tokens
Context length 1M, max output 384K, model version DeepSeek-V4-Pro-0813
Concurrency limit 500
Per the vendor's own announcement, API service continues past September 14, 2026 at unchanged pricing

Current models

Short overview of the vendor's current main models. The API calculator may list more snapshots and variants.

DeepSeek-V4.1-Flash

Lower-priced variant with a higher concurrency limit; now unified with image input.

  • 1M-token context, up to 384K output; thinking and non-thinking modes; supports image input.
  • API concurrency 2500 per the pricing page; peak and off-peak rates. Supersedes the earlier V4-Flash naming (legacy names deepseek-v4-flash / deepseek-v4-flash-vision-exp still work but route here).

Source https://api-docs.deepseek.com/quick_start/pricing/

DeepSeek-V4-Pro

Higher-priced V4 variant; web chat lists access to DeepSeek-V4-Pro.

  • Same context and feature table as Flash; API concurrency 500.
  • Doc examples often use deepseek-v4-pro with thinking set to high; the vendor continues API service past September 14, 2026 at unchanged pricing.

Source https://api-docs.deepseek.com/quick_start/pricing/

Limits

The currently checked vendor page does not fully expose reliable price or usage limits. Earlier amounts are therefore not presented as current vendor prices.

Similar tools

From the same subcategory “Assistants & workspaces”, continuing alphabetically from this entry.

DeutschlandGPTGDPR-oriented interface for multiple AI models (ChatGPT, Claude, Gemini, Llama, Mistral) with German hosting.FreeDiaAI-native web browser from The Browser Company that embeds an AI assistant with access to open tabs and connected work tools directly in the browser.Better AnswersDolaEveryday chat assistant from SPRING (Singapore), formerly known as Cici.DolaDoubaoMultimodal AI assistant by ByteDance for text, image, and voice interactions, primarily offered for the Chinese market.DoubaoDuck.ai (DuckDuckGo)Privacy-oriented multi-model chat with a free daily limit and higher limits in DuckDuckGo Plus and Pro.DuckDuckGo PlusEntergramTelegram CRM with a shared inbox, team workspaces, and MCP access for AI clients.BasicERNIE (Baidu)Baidu's ERNIE model family, available as a Chinese-language chat assistant at wenxin.baidu.com and as a paid API through the Baidu Qianfan platform.Qianfan International API - ERNIE 5.0 (Text generation, usage-based)FanarArabic-language AI platform by QCRI/HBKU (Qatar) for chat, image generation, and culturally aware responses.Fanar

FAQ

Is DeepSeek available for free?

Yes. A permanently free plan is published, not a time-limited trial.

Where is DeepSeek's vendor based?

Hangzhou DeepSeek Artificial Intelligence Co., Ltd., based in China. For data processing and privacy terms the vendor's own contract governs.

Is DeepSeek still active?

Active under the name DeepSeek. The vendor (Hangzhou DeepSeek Artificial Intelligence Co., Ltd.) lists DeepSeek Web, DeepSeek App, DeepSeek Harness and API Platform on its home page. Current flagship per the home page: DeepSeek-V4-Pro, available on web, app and API.

How many plans are documented for DeepSeek?

The catalog documents 3 plans with price, billing period, and included features.

Which use case is DeepSeek categorized for?

DeepSeek is listed in the catalog under “Productivity.”

Where do the details about DeepSeek come from?

Prices, limits, and product details are documented against the listed official vendor page.

How we compare

We compare free access, paid entry, plan limits, features and the listed official sources. Unsourced figures stay marked as not published.

Only confirmed entries appear in the catalog. Prices, plans and limits come from listed official vendor pages. Missing figures stay marked as not published. Result lists are alphabetical by default and can be switched to Free access first; neither is a quality ranking. The vendor page always wins.

Edited and checked by the AI Compare team

Sources

Visit site (opens in a new tab)

Spotted an outdated figure? Vendor pages change without notice. Report an error