DeepSeek’s Open-Source AI Agent Harness Hits 1.54 Million Daily Users

DeepSeek’s AI Agent Platform Sees Rapid Growth

DeepSeek is gaining attention in the artificial intelligence industry once again. Its open-source AI agent application, DeepSeek Harness, has reportedly reached 1.54 million daily active users in less than three months after its launch.

The company also revealed that Harness users consume approximately 92 trillion tokens every day. These figures highlight the growing demand for AI agents that can perform complex tasks rather than simply answer questions.

DeepSeek shared the usage numbers on November 8, 2026, with the user figure recorded as of November 4. The announcement marks an important milestone for the company’s growing AI agent ecosystem.

The figures cover users connecting through DeepSeek’s official API. They do not include every person using self-hosted versions or accessing the software through other services.

What Is DeepSeek Harness?

DeepSeek Harness is an open-source application designed to turn AI models into agents that can perform multi-step tasks.

Traditional AI chatbots mainly respond to questions and instructions. AI agents can go further by using tools, interacting with software, analysing files and carrying out a sequence of actions to complete a task.

For example, a user could ask an AI agent to research a topic, collect information from different sources, organise the findings and prepare a report.

In software development, an agent could inspect a codebase, identify errors, modify files and run tests with limited human involvement.

Harness provides the infrastructure needed to manage these activities. It connects AI models with tools, plugins, sessions and other capabilities that allow them to complete tasks.

The application is built around an open and flexible architecture, giving developers the ability to customise its components.

1.54 Million Daily Users in Less Than Three Months

DeepSeek released the first version of Harness on August 13, 2026. By November 4, the application had reached 1.54 million daily active users through its official API.

The company had reported around 271,000 daily users at the end of September. This suggests substantial growth in a relatively short period, although the earlier and latest figures were reported using different measurement periods.

The desktop application, released on September 29, may also have helped increase adoption by making the platform easier to access.

The reported growth shows that interest in AI agents is extending beyond traditional chatbot applications.

Developers and other users are increasingly exploring tools that can manage longer workflows, interact with external applications and automate everyday activities.

92 Trillion Tokens Every Day

One of the most notable figures in DeepSeek’s announcement is the reported daily token consumption.

Harness users process approximately 92 trillion tokens per day, according to the company.

Tokens are small units of text or other information that AI models process. They can represent complete words, parts of words or other pieces of information.

The high consumption reflects the way AI agents operate. Unlike a chatbot that may respond to a single question, an agent can make repeated model calls while completing a task.

For example, an agent working on a software project might need to:

• Read and understand existing code.
• Search for relevant files and information.
• Plan the required changes.
• Generate or modify code.
• Run tests and review results.
• Correct errors and repeat the process.

Each stage can require additional model processing. Longer tasks and repeated interactions can therefore consume significantly more tokens.

The reported 92 trillion figure also offers an indication of the computing resources required to support large-scale AI agent applications.

However, token consumption alone does not reveal how many tasks were completed successfully or how much revenue the platform generates.

A Growing Ecosystem With 35,000 Plugins

DeepSeek is also building an ecosystem around Harness through third-party plugins.

The company reported approximately 35,000 plugins developed by the community. It also said that around 60% of active users had created or installed at least one third-party plugin.

Plugins allow developers to add new capabilities without changing the entire application.

For example, a plugin could help an agent work with a particular software tool, organise research materials, generate content or automate a specialised workflow.

This approach gives users more flexibility and allows developers to create solutions for specific needs.

DeepSeek has also indicated plans to introduce an official plugin marketplace. The marketplace is expected to make it easier for developers to publish their plugins and for users to discover useful additions.

The company is also preparing a support programme for plugin creators, with possible token incentives, technical assistance and promotional opportunities.

The marketplace could help turn Harness into a larger platform where third-party developers contribute new tools and services.

Why DeepSeek’s Growth Matters for the AI Industry

The growth of DeepSeek Harness reflects a wider shift in artificial intelligence.

The competition is no longer limited to building powerful language models. Technology companies are also competing to create platforms that help users apply AI to real tasks.

AI agents could become useful in areas such as:

• Software development: Writing code, debugging and testing applications.
• Research: Collecting information, reviewing documents and preparing reports.
• Business operations: Automating repetitive processes and managing workflows.
• Data analysis: Processing files and identifying useful information.
• Customer support: Handling multi-step requests and connecting with business systems.

Open-source platforms could make these capabilities more accessible to developers and organisations.

DeepSeek’s plugin-based approach also gives the wider developer community an opportunity to extend the platform beyond its original features.

However, the quality, security and reliability of third-party plugins will remain important considerations as the ecosystem grows.

What Comes Next for DeepSeek Harness?

DeepSeek’s next major step could be the launch of its official plugin marketplace and developer support programme.

These initiatives may help attract more developers and expand the range of tasks Harness can perform.

The company will also need to address practical challenges involving API stability, compatibility, security and the quality of community-developed plugins.

Another important question is how DeepSeek will sustain the infrastructure required to process such a large volume of tokens.

The reported numbers are significant, but they are company-reported figures and do not provide a complete picture of revenue, paid usage or long-term user retention.

AI Agents Are Moving Beyond Chatbots

DeepSeek Harness reaching 1.54 million daily users in less than three months highlights the growing interest in autonomous AI applications.

Its reported 92 trillion daily tokens and 35,000 third-party plugins also show how much infrastructure and developer participation may be required to support this new category of AI tools.

As companies continue developing more capable agents, the focus will increasingly move towards practical applications, integrations and complete workflows.

DeepSeek’s progress suggests that the next stage of AI competition may be shaped not only by the models themselves, but also by the platforms and ecosystems built around them.