DeepSeek Prices Web Search at 75 Cents per 1,000 Queries, Undercutting OpenAI and Anthropic

DeepSeek is charging developers 75 cents per 1,000 web searches for the search engine behind its chatbot. That is less than a tenth of the $10 per 1,000 that OpenAI and Anthropic list for the search tools in their own APIs.

The price arrived in a release published on Friday, November 13, 2026, that moves nearly everything the DeepSeek App does into the company's API. The new capabilities include cited web search, a multistep "deep thinking" search, file parsing, link reading, content safety checks, OCR and speech in both directions. Announcing the launch on X, the company wrote: "The full DeepSeek App experience, now via API. Search, files, links, safety, and voice. The capabilities behind the DeepSeek App are now available through the DeepSeek API."

The release changes how developers build on DeepSeek. Before Friday, developers who wanted those features had to bolt outside tools onto DeepSeek's text model and handle the hand-offs themselves, as Chinese outlet The Paper noted. Now one provider can cover the whole chat stack on a single bill.

What does the release actually include?

The built-in features run on deepseek-flash, the API name for the V4.1-Flash-1017 model. They work across the Messages, Responses and Chat Completions formats "with tool calls and context assembly handled server-side," DeepSeek said.

Activation works through declared parameters rather than separate plumbing:

• A developer turns on search by declaring a deepseek_web_search tool.
• Link reading is enabled with deepseek_read_url.
• With thinking mode on, the same search tool becomes deep search, and the model can make up to about 20 tool calls before it answers, according to the company's web tools guide.

For pipelines where code, not the model, decides when to call a tool, DeepSeek also split out standalone Search, Fetch, OCR, speech-to-text and text-to-speech endpoints.

How much does each piece cost?

The pricing undercuts U.S. rivals on most line items, though not all of them. OCR costs $0.225 per 1,000 pages. Both speech services are half off for a limited time, which puts transcription at $0.058 per hour of audio and synthesis at $0.225 per 10,000 characters. DeepSeek has not said when the speech discount ends; its pricing page states only that the company "reserves the right to adjust" prices.

The speech endpoints come with constraints. The speech-to-text endpoint detects 21 languages but takes clips of no more than 120 seconds and does not stream, according to its documentation. Text-to-speech offers two voices in 10 languages and can stream over WebSocket.

Token prices sit at the low end of the market as well. On DeepSeek's pricing page, deepseek-flash runs $0.30 per million uncached input tokens and $1.20 per million output tokens at peak, and half that off-peak. The peak windows are set in UTC and fall overnight in the United States — from 8 to 11 p.m. and 1 to 5 a.m. Eastern time — so American daytime traffic gets the lower rate.

Where does the price advantage stop?

The headline gap narrows on fetch operations. OpenAI's pricing page lists web search at $10 per 1,000 calls, with search content tokens billed at model rates. Anthropic charges $10 per 1,000 searches plus token costs, though its web fetch tool carries no fee beyond tokens, according to Claude's pricing documentation.

DeepSeek bills fetches at the same 75 cents per 1,000 as searches, whether the developer calls them directly or the model does. That structure produces one inversion: an agent that reads far more pages than it searches could pay more for fetching on DeepSeek than on Claude.

How large is the underlying index?

The company is also pitching the quality of its search infrastructure, not just its price. In its change log, DeepSeek said its in-house engine indexes hundreds of billions of web pages and handles hundreds of millions of searches a day. The engine "filters SEO spam, GEO manipulation, and low-quality AIGC through strict detection mechanisms," the company said. Those are DeepSeek's own figures and have not been independently verified.

What did DeepSeek open-source?

To show the pieces working together, DeepSeek open-sourced chat-demo, an MIT-licensed web client available on GitHub. The client has no backend; it connects from the browser straight to the API and can import a user's DeepSeek chat history. The user pastes an API key into its settings before the first chat.

What strings are attached?

The release carries conditions that will shape who can adopt it. Public-facing apps must turn on the safety checks, and developers "remain responsible for content moderation, applicable model registration, and AI-generated content labeling," DeepSeek said.

U.S. public-sector buyers face a separate limit. A Department of Homeland Security order requires civilian federal agencies to remove DeepSeek products and services built on its models. That order wallsoff one segment of the market even as DeepSeek courts developers elsewhere.

The release also lands at a specific corporate moment. DeepSeek hired its first chief financial officer in September 2026 as part of a push toward a Shanghai initial public offering, and the company continues to work toward a possible Shanghai listing. A broad, competitively priced API business would strengthen that story for prospective investors.

Why does the pricing gap matter?

The 75-cent price point compresses the economics of search-augmented AI applications. A developer running one million searches a month would pay $750 to DeepSeek for the search calls alone, against $10,000 at OpenAI's or Anthropic's list prices before token charges. That arithmetic gives Chinese-model-based agents a structural cost advantage in any product where retrieval is the dominant expense, and it puts pressure on U.S. providers' search-tool margins even before any matched price cut.

The bundled stack also removes a category of integration work. Developers who previously assembled search, document parsing, OCR and speech from separate vendors now have a single vendor contract, one authentication layer and one bill for the full chat experience — the same building blocks DeepSeek ships in its own consumer app.

Whether rivals respond with price cuts or counter with capability arguments, DeepSeek has reset the reference price for API web search at under a dollar per thousand queries.