Deployment-free, 0-cost use of Deep Research Web UI

Deep Research Web UI is an intelligent research visualization tool driven by AI. It integrates advanced technologies such as search engines, web crawling and large language modeling to automatically deep-dive into complex issues and generate well-structured research reports. The tool emphasizes users' data security and deployment flexibility, with all data processing done in a local browser and support for private deployment. Users can track the logic of the study in real time through the dynamic tree structure, and the final report supports one-click export to PDF or Markdown format for easy sharing and archiving.

Currently.Deep Research Web UI Already support Silicon-based flow Big Model API services provided by the SiliconCloud platform. Users have the flexibility to choose the SiliconCloud platform includes DeepSeek-R1 The SiliconCloud platform is designed to meet a wide range of AI big models such as V3 and V3 to meet a wider range of research needs and application scenarios, and to experience more efficient, safer and more flexible intelligent research services. The SiliconCloud platform is known for its high-performance and cost-effective big model services, especially the DeepSeek series of models that perform well on a variety of evaluation benchmarks, providing users with powerful AI computing power support.

In this article, we will introduce in detail how to configure and access the API services provided by Silicon Mobility in Deep Research Web UI to help users get started quickly and improve research efficiency.

1. Configuring the Big Model API

First, open the official Deep Research Web UI website (https://deep-research.ataw.top/). In the top right corner of the page, click on the gear-shaped "⚙️" settings button to open the settings pop-up window.

Configuration Model Service Provider

In the Settings pop-up window, find the AI Big Model Service Configuration section. In the Model Service Provider drop-down menu, select SiliconFlow SiliconFlow.

Getting and Configuring API Keys

Next, you will need to obtain an API key for SiliconCloud. Visit the SiliconCloud website (https://cloud.siliconflow.cn/models) and register or login to your account. After logging in, generate and copy a new API Key on the API Key Management page. Please keep your API Key in a safe place to avoid leakage.

Return to the Settings pop-up on the Deep Research Web UI page. In the AI Service Configuration section, make sure "SiliconFlow Silicon Flow" is selected. In the "API Key" input box, paste the API Key you just copied.

Select model name and configuration context length (optional)

In the Model Name column, click the drop-down button on the right, or enter a model name to filter and select the model you wish to use. Deep Research Web UI supports a variety of models on the SiliconCloud platform, including DeepSeek-R1 and V3. Users can choose according to their own research needs and preferences.

It is recommended to configure the "Context Length" option if you need to conduct large-scale or research that requires processing long text. The context length determines the maximum amount of text that the model can process in a single request. Please note that the context length set should not exceed the maximum context length supported by the selected model to avoid request failures.

2. Configuring the networking search model

The Deep Research Web UI currently supports two internet search services, Tavily and Firecrawl, with plans to support more in the future. Tavily offers 1000 free searches per month, which is sufficient for most users. Tavily offers 1000 free searches per month, which is sufficient for most users, while Firecrawl may focus on data capture and coverage, allowing users to choose the right search service for their needs.

Get and configure Tavily API key

To use the Tavily Networked Search Service, first visit the official Tavily website (https://app.tavily.com/home) to register for an account. Once logged in, go to the console to create a new API Key and copy it.

When creating an API Key, you can customize the Key Name, for example "Deep Research". The "Key Type" can be selected according to your usage frequency, select "Development" for light usage and "Production for light use and "Production" for heavy use. Please make sure to keep your API Key in a safe place to prevent leakage.

Configuring the Deep Research Web UI Networked Search Service

Go back to the Settings pop-up on the Deep Research Web UI page and locate the "Networked Search Services" section. In the Service Provider selection, select "Tavily". In the "API Key" input box, fill in the API Key you just generated in the Tavily console.

Setting the search language (optional)

You can set the language used for searching based on your research needs. By default, the AI model searches and responds in the language of your current web page. If you would like to use English search terms for higher quality English material, you can set the "Use language" setting to "English". English search results typically have broader and deeper coverage in certain technical areas and academic research.

Once the above configuration is complete, the Deep Research Web UI is ready to connect to the SiliconCloud Big Model API for Silicon Flow and the Tavily networked search service. Click Save Settings to get started.

3. Getting started with the Deep Research Web UI

The Deep Research Web UI is designed to be low-barrier for users, with every step of configuration made simple and straightforward. Once configured, you can utilize it for a variety of research tasks, such as:

• Find first-hand information on. Quickly retrieve and analyze the latest research papers, essays, and industry information on the web.
• Learn about topics of interest. Digging deeper into specific areas or concepts to build a body of knowledge.
• News event tracking and summaries. Collects and organizes relevant stories and commentary on specific news events into thematic reports.

For example, you can try searching for "information on NVIDIA RTX 50 series graphics cards", and the Deep Research Web UI will automatically perform deep digging and analysis, and visualize the research vein in a tree structure, and ultimately generate a detailed report on the card.

The Deep Research Web UI project is under continuous active update, if you encounter any problems or have suggestions in the process of using it, please go to the GitHub repository (https://github.com/AnotiaWang/deep-research-web-ui) for feedback. The developer community welcomes user participation and contributions to improve this AI-driven intelligent research tool.