DeepSeek lets developers plug its search and file-reading features into their own apps
DeepSeek's app features, like web search and file reading, can now be called by other software, with extra charges on top of normal model fees.
The short version
- Announced the evening of November 13: web search, deep-thinking search, file input and web page reading join the existing interface.
- Developers switch on the matching settings to use these features in their own apps.
- Text recognition, speech recognition and speech synthesis are available through separate interfaces.
- The chat features currently run on the V4.1-Flash model.
- Tool calls cost extra on top of the model's token fees.
Why it matters
Features that used to be reachable mainly through DeepSeek's own app can now be built into other products. If you use small apps or services, some may start offering DeepSeek-powered search or document reading. For anyone building with it, the separate tool charges mean the bill is more than the model fee.
What DeepSeek announced
As Phoenix New Media (ifeng.com) reported, DeepSeek, a Chinese AI lab, said on the evening of November 13 that it is opening its application features through its API. An API is a doorway that lets one piece of software use another's service directly.
Four abilities are being folded into DeepSeek's existing interface: web search, deep-thinking search, file input and web page reading. Until now, most people met these through DeepSeek's own app.
How developers switch it on
According to the report, a developer turns on the matching setting in their request. After that, their own app can call the same ability, so a note-taking tool, for example, could search the web or read an uploaded file without building that from scratch.
Some other functions are handled separately. Text recognition (turning images of text into editable text), speech recognition and speech synthesis can each be called through their own interfaces.
What it costs
The chat features are currently powered by a model called V4.1-Flash. Pricing has two layers. Developers pay the usual model fee, which is counted in tokens (small chunks of text the model reads and writes). Tool calls, such as a search or a page read, are charged on top.
The source gives no prices, usage limits or complete list of supported tools, so any cost estimate would be a guess.
What to check before using it
Teams considering this should read DeepSeek's own documentation and price list first. Search and file reading can involve customer or company data, so it also matters where that data goes and how it is stored. The announcement as reported does not say.
What to watch
- What the separate tool-call charges and usage limits turn out to be.
- Whether DeepSeek explains how data sent through these tools is handled and stored.
Try it
If you or a developer colleague use DeepSeek's API, look up its price list and compare the tool-call charges with the model's token fees before building search or file reading into a product.
Sources and links
Liked it? Say why
Add a line for other readers.
No account needed. We never show your email. Comments with links, ads or abuse are held and not shown. Your comment shows after a quick check.
Listen to the AI Wire headlines
Listen to the AI WireListen to our daily AI shows every morning: on your commute, in the shower, wherever you are
Listen to the daily showsFirst reported by Phoenix New Media (ifeng.com) · Read the original. Written by the SI Agent News Desk from the reporting above.Written with AI from the source above. Every number is checked against the source. Spot a mistake? Corrections.
