TEN stands for Transformative Extensions Network and represents the Next-Gen AI-Agent Framework, the world's first truly real-time multimodal AI agent framework.

Stay Tuned

Before we get started, be sure to star our repository and get instant notifications for all new releases!

Features

The TEN framework offers the following advantages:

1. Native Support for High-Performance, Real-Time Multimodal Interactions

   If your AI applications involve complex audio-visual scenarios, TEN is your go-to solution. It offers high performance and low latency, with extensive optimization of interactions between various extensions to ensure efficient development of AI applications.

2. Supports Multiple Languages and Platforms

   Developers can create modular and reusable extensions using various programming languages, such as C++, Go, and Python (with future support for JavaScript/TypeScript). Moreover, the TEN framework runs seamlessly across platforms, including Windows, Mac, Linux, and mobile devices.

3. Edge-Cloud Integration

   Through the TEN framework, extensions deployed across edge and cloud environments can be easily combined to create diverse applications and scenarios. For privacy-sensitive edge deployments, small models leverage local compute power for reduced costs and lower latency, while cloud-based large models can be integrated for an optimal balance of cost and performance.

4. Flexibility Beyond Model Limitations

   The TEN framework allows for the creation of complex AI applications that transcend the limitations of large models alone. Agents can be easily constructed to meet a wide range of needs using a simple drag-and-drop, responsive programming approach. TEN also facilitates the integration of AI with audio-visual tools, databases, monitoring systems, RAG, and more.

5. Real-Time Agent State Management

   TEN has the ability to manage real-time agent states, enabling dynamic responsiveness and adjustment of agent behavior in real time.

6. And more...

   For more information and detailed documentation on all the capabilities of the TEN framework, please refer to the TEN framework documentation site.

Use Cases

With the TEN framework, you can easily accomplish the following scenarios. You can see actual demos at TEN Agent:

1. Simultaneous interpretation

   Real-time language translation during live conversations, enabling smooth cross-language communication without delays.

2. Speech-to-text conversion

   Convert spoken language into written text, making it useful for transcribing meetings, interviews, or live talks.

3. Multilingual chat rooms

   Create chat rooms where users can communicate in different languages, with automatically translating messages in real time to foster seamless interaction.

4. Audio interaction

   Enabling users to communicate with the AI using audio instead of text, which is ideal for hands-free communication or enhancing accessibility.

5. Audio-visual interaction

   Combine audio and visual elements to create interactive experiences, such as video conferences with integrated real-time transcription, translation, or even interactive media content.

6. And more...

   The flexibility of the TEN framework enables developers to build additional interactive solutions, such as virtual assistants, automated customer support systems, and dynamic multimedia applications.

Documentation

Please visit our TEN framework documentation site for more information.

License Information

1. The entire TEN framework (except for the folders explicitly listed below) is released under the Apache License, Version 2.0, with additional restrictions. For details, please refer to the LICENSE file located in the root directory of the TEN framework.

2. The components within the packages directory are released under the Apache License, Version 2.0. For details, please refer to the LICENSE file located in each package's root directory.

3. The third-party libraries used by the TEN framework are listed and described in detail. For more information, please refer to the dependencies.md file located in the docs/ten_framework directory.