AgenticSeek: A Local Manus Alternative with Voice Input Support 206

What is AgenticSeek?

AgenticSeek is an open-source, locally-running AI assistant, aiming to be an alternative to Manus AI. It is powered by the Deepseek R1 model and runs entirely on local hardware, with no dependency on cloud services. This means users don't have to worry about data privacy issues or pay expensive cloud service fees.

Main Features of AgenticSeek

• 100% local execution: All operations are done locally; data is not uploaded to the cloud.

• Voice interaction: Supports voice input and interaction.

• File system operations: Can manipulate local files via Bash commands.

• Code writing and debugging: Supports multiple programming languages (such as Python, C, Golang, etc.), and can write, debug, and run code.

• Autonomy: If a command fails or the code errors out, it automatically retries and fixes it.

• Smart routing: Automatically selects the most suitable AI agent for the task.

• Task decomposition: For complex tasks, it launches multiple agents to collaborate.

• Tool support: Each agent is equipped with a variety of tools (such as search, file exploration, etc.).

• Memory: Able to remember useful information, user preferences, and historical conversations.

• Web browsing: Supports autonomous web browsing.

Installation and Running of AgenticSeek

• Hardware requirements: A Deepseek model of at least 14B is recommended, requiring at least 12GB of GPU VRAM (e.g., RTX 3060).

Installation steps:

• Clone the repository and set up the environment.

• Create a Python virtual environment and install dependencies.

• Use Ollama to download and run the Deepseek model.

• Configure the config.ini file to specify the model and run mode.

Running modes:

• Can be run locally, or the model can be run on a remote server and accessed over the network.

• Supports using external services via API (such as OpenAI, HuggingFace, etc.).

AgenticSeek Applications:

• Programming and Bash: For example, helping users complete matrix multiplication code (Golang), scan the network (nmap), write a Snake game (Python), etc.

• Web search: For example, searching for Japanese AI startups, finding the creator of agenticSeek, etc.

• File system operations: For example, finding lost files, checking disk space, etc.

• Daily conversation: For example, telling jokes, querying flight information, discussing the meaning of life, etc.

GitHub Repository:https://github.com/Fosowl/agenticSeek