Free Intensive Nov 21–22Free 2-day hands-on Vibecoding Intensive on Nov 21–22 at 16:00 MSK (Sat–Sun) · Live Cohort 3 starts Nov 23Join Free →Cohort 3 (Nov 23)

DeepSeek Opens API Access to All App Features: Search, Files, Voice

DeepSeek has opened up through its API almost everything that was previously available only in its app: web search, deep thinking mode, file handling, and link reading. Voice is also included, along with OCR for text in images, ASR for speech recognition, and TTS for speech synthesis. The company has also published a demo chat frontend that can import history from the app. For teams, this means the app can now be recreated almost one-to-one without assembling it from third-party services.

AI-processed from TG: @bezsmuzi (Русский ИТ бизнес); edited by Hamidun News
DeepSeek Opens API Access to All App Features: Search, Files, Voice
Source: TG: @bezsmuzi (Русский ИТ бизнес). Collage: Hamidun News.
Listen to article
Article Fact Sheet & Key Figures
ParameterDetails
Primary sourceTG: @bezsmuzi (Русский ИТ бизнес) ↗
Editor-in-Chief
Key metric #1Web search, deep thinking, files, and link reading are available through the API
Key metric #2Voice, OCR, speech recognition, and speech synthesis are already part of the package
Key metric #3The demo chat frontend is open and can import history from DeepSeek

Previously, a developer who wanted to build "their own ChatGPT" on the DeepSeek model ran into one problem. The model itself is available through the API, but everything users love about the app had to be assembled by hand. That includes internet search, document uploads, link parsing, and voice mode. Now the company has closed that gap: according to its announcement, almost the entire feature set of the official app is now open through the API.

What Is Now Available Through the API

According to the announcement, developers now have access to web search, deep thinking mode, file processing, and link reading. These are the features that turn a chatbot from a "conversation partner with training-data memory" into a working tool. The bot can go out to the web for fresh data, read a PDF or spreadsheet, open an article at a given address, and parse it.

Voice interaction is mentioned separately, along with three foundational technologies: OCR, ASR, and TTS. OCR recognizes text in photos and scans, ASR converts speech into text, and TTS reads the response aloud. Together they make a complete voice loop: the user speaks, the system listens, thinks, and responds by voice. Previously, this usually required connecting two or three third-party services, each a separate source of latency, billing, and errors.

Demo Frontend and History Import

Perhaps the most practical part of the news is not the feature list but the fact that the company has released a ready-made demo chat frontend. It is a working interface that can be used as a starting point rather than drawn from scratch. It supports importing conversation history from DeepSeek.

For a product, this is an important detail. Suppose a team is building an enterprise assistant on DeepSeek. Employees have already used the public app, and they have accumulated dialogues with working templates. Moving that history into a custom interface lowers the barrier to entry: people don't lose context and don't start from an empty screen. According to the author of the post, developers can now practically recreate the DeepSeek app one-to-one.

What This Changes for Developers and Product Managers

First: the cost of building a prototype drops. If search, files, and voice come from a single API, there's no need to build a pipeline of search results, parsers, and speech recognition yourself. A minimum viable product can be delivered faster, and the team can focus on use cases rather than integrations.

Second: fewer scattered vendors. Each additional service in the chain means another contract, another limit, and another point of failure. A unified feature set from one provider simplifies operations.

Third: dependence on a single vendor grows. If the entire product, from search to voice, rests on the DeepSeek API, then changing the model or the terms would require rewriting a significant part of the system. Those building a serious service should sensibly plan an abstraction layer from the start, so individual components can be replaced when needed.

What the Announcement Doesn't Say

The source, a Telegram channel, retells the announcement very briefly, so many important details are left out. The post contains no pricing for the new features, no request limits, no list of supported file formats, and no list of voice languages. It doesn't say how closely the quality of web search and OCR through the API matches what users see in the app. The phrase "almost all capabilities" also means some features remain only in the official client, but the author doesn't specify which ones.

Therefore, before moving a product onto this, it's worth checking the DeepSeek documentation: pricing terms, file size limits, how voice mode behaves under load, and the data retention policy. The last point matters especially for companies planning to send internal documents through the API.

Bottom Line

DeepSeek is moving from the role of a "model supplier" toward the role of a platform on which a full consumer product can be built. For small teams and solo developers, this is a convenient way to quickly get a working chat with search, files, and voice. But the figures on price and quality will have to be found in official materials, not in the announcement.

Story

Key Questions Answered

What Is Now Available Through the API?

According to the announcement, developers now have access to web search, deep thinking mode, file processing, and link reading. These are the features that turn a chatbot from a "conversation partner with training-data memory" into a working tool. The bot can go out to the web for fresh data, read a PDF or spreadsheet, open an article at a given address, and parse it. Voice interaction is mentioned separately, along with three foundational technologies: OCR, ASR, and TTS.

What are the key details on Demo Frontend and History Import?

Perhaps the most practical part of the news is not the feature list but the fact that the company has released a ready-made demo chat frontend. It is a working interface that can be used as a starting point rather than drawn from scratch. It supports importing conversation history from DeepSeek. For a product, this is an important detail. Suppose a team is building an enterprise assistant on DeepSeek. Employees have already used the public app, and they have accumulated dialogues with working templates.

What This Changes for Developers and Product Managers?

First: the cost of building a prototype drops. If search, files, and voice come from a single API, there's no need to build a pipeline of search results, parsers, and speech recognition yourself. A minimum viable product can be delivered faster, and the team can focus on use cases rather than integrations. Second: fewer scattered vendors. Each additional service in the chain means another contract, another limit, and another point of failure. A unified feature set from one provider simplifies operations.

ZK
Zhemal Khamidun — CPO at Alpina Digital, product manager of AlpinaGPT. Building AI products since 2019. Hamidun News is a pet project that grew into a production aggregator with 6,000+ readers.
✦ Unified AI Workspace · Nomad

Test Frontier LLMs & Autonomous Agents in One Workspace

Run Claude Opus 5.5, GPT-6, Gemini 3.1, and DeepSeek side-by-side in Nomad AI (nomadnet.ai) — unified model routing, deep research, and multimodal workflows under one subscription.

Try Nomad AI →Launch Nomad Agent
What do you think?
Loading comments…