Skip to main content

A distributed, agentic AI platform — run AI agents everywhere, on every device.

Project description

Subconscious — Your Distributed AI Agent

Subconscious Logo

A distributed, agentic AI platform that allows you to create AI agents running everywhere, on all your connected devices.
An open-source alternative to ChatGPT, Claude & Grok; A packaged alternative to Open Claw & Open WebUI

Build and Release GitHub Stars PyPI Version Documentation

Subconscious UI Demo

✨ Key Features

  • 🔒 Local-First Architecture: Your data is exclusively yours. All chats, logs, and settings are stored locally.
  • 🖥️ Run Local Models: Execute models completely offline using powerful local servers like Ollama.
  • 🔑 Bring Your Own Keys (BYOK): Use your own API keys to connect to the models of your choice.
  • 🛠️ Built-in Tools Out-of-the-Box: Outfitted with numerous handy local tools. Just ask your Subconscious what it can do! Manage access natively inside settings. Currently supports: Terminal, FileSystem, Calculator, Clipboard, Contacts, Time, Weather, Web Tools, and more!
  • 📱 True Multi-Platform Support: Works beautifully via Python and Windows executables out-of-the-box.
  • 🔄 Auto-Updates: Subconscious automatically checks for the latest features, with a convenient one-button seamless update.
  • 🧵 Agentic Threads: Create varied threads/chats/agents for distinct purposes; fine-tune context effortlessly.
  • 📁 Workspaces: Keep organized. Compartmentalize your threads by workspace for related themes, such as 'Work', 'Personal', or a special side-project.
  • ⏱️ To-Dos & Scheduled Tasks: Delegate routine chores. Let your agent maintain a to-do list or schedule tasks for exact execution times.
Sidebar & Workspaces Screenshot

🚀 Getting Started Flow

Subconscious is designed to be as ubiquitous as you are. Use your phone to execute a process on your VPS, or operate Subconscious in your terminal while browsing.

Installation (Recommended)

Microsoft Windows (via Winget)

The fastest way to install the Subconscious Desktop App natively on Windows:

winget install Ancilla.Subconscious-Chat

Via Python Pip

For Mac, Linux, and Windows users comfortable with Python:

Create and activate a new Python environment (e.g., using venv or conda), then install via PyPI:

pip install -U subconscious-chat

Note: If using pip, you will need Python 3.12 or newer.

🛠️ From Source (Development)

Love tinkering? Getting Subconscious running locally via source is easy.

git clone https://github.com/Ancilla-Company/Subconscious.git
cd Subconscious

# Alternatively, use the included VS Code Task "Pip Install Editable"

pip install -e .
Settings & Model Configuration Screenshot

Supported Models

With our native integration to the best APIs and local servers, Subconscious flexibly connects to:

  • OpenAI (GPT-4o, etc)
  • Anthropic (Claude 3.5 Sonnet/Opus, etc)
  • Google GenAI (Gemini)
  • Ollama (Llama-3, Phi, Mistral — 100% Offline)
  • DeepSeek
  • Mistral
  • Groq
  • Grok
  • Hugging Face

Running Subconscious

Desktop App

If you installed the Winget package or built the binary via Nuikta/PyInstaller, simply launch Subconscious from your Start menu / App launcher!

CLI / Python

Once installed via pip, start the agent directly:

subconscious

🗺️ Roadmap & Highlights

We are shipping fast. Here is a glimpse of what's currently being integrated:

  • Swarm Access: Start on one device and continue from another. Begin a task in your browser and seamlessly transition to your phone.
  • Agentic Threads: Create varied threads/chats/agents for distinct purposes. Assign specific tools and fine-tune context effortlessly.
  • True Multi-Platform Support: Web, Linux, MacOS, Browser, Android and iOS are actively coming soon!
  • Custom & External Tools: Built in Python natively, meaning extending Subconscious with your own tools is a breeze. Connect to external APIs or services to supercharge your workflow.
  • More Models Coming Soon: We are actively working on expanding support for as many models as possible.
  • Voice Input Button: Speak directly to Subconscious with accompanying outputs/animations.
  • Computer Interaction Tools: Allow the AI to control mouse/keyboard using computer usage agent protocols.
  • Directory RAG System: Instantly query your local filesystem docs.
  • Mini Mode / Always-On-Top: Perfect for persistent UI integration while researching.
  • MCP Server Connections: Next-gen integrations for extended memory and actions.

🤝 Contributions

Subconscious is open source and community-driven! We welcome bug reports, pull requests, and ideas. Please check out the issues tab or submit a PR for a feature from the roadmap.

📜 License

This project is open-source and available to the community under a custom license based on BSD-3-Clause with additional branding restrictions.

Third-Party Components

Subconscious bundles various third-party libraries. See THIRD_PARTY_LICENSES.txt for a complete list of bundled components and their licenses.

LGPL Components

Subconscious includes LGPL-licensed components (notably pystray for system tray functionality). In compliance with LGPL requirements:

  • The source code for LGPL components is available at their respective upstream repositories
  • Users may modify and update these components independently
  • For questions about LGPL compliance, please contact the maintainers

Enterprise Licensing

For commercial/enterprise use beyond the open-source license terms, please contact the maintainers for enterprise licensing options.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

subconscious_chat-0.1.7.tar.gz (369.5 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

subconscious_chat-0.1.7-py3-none-any.whl (365.5 kB view details)

Uploaded Python 3

File details

Details for the file subconscious_chat-0.1.7.tar.gz.

File metadata

  • Download URL: subconscious_chat-0.1.7.tar.gz
  • Upload date:
  • Size: 369.5 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for subconscious_chat-0.1.7.tar.gz
Algorithm Hash digest
SHA256 b066ed05fb26a4fd80ad5ea57efa6d8f9134879eb6ef88971742d56caeaedc35
MD5 5905d0d621b093e1c0d8d413db424c64
BLAKE2b-256 7af42ebb0a23f02d749d6c038e37a1750735fd8563822955069c6d200f685b89

See more details on using hashes here.

Provenance

The following attestation bundles were made for subconscious_chat-0.1.7.tar.gz:

Publisher: release.yaml on Ancilla-Company/Subconscious

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file subconscious_chat-0.1.7-py3-none-any.whl.

File metadata

File hashes

Hashes for subconscious_chat-0.1.7-py3-none-any.whl
Algorithm Hash digest
SHA256 c73325caf88445fcb8d97c8b72a45a7bf7a702c3bc9e20ccb0b578c07df817fe
MD5 1bceacd752eb67af161a9d701fb403bd
BLAKE2b-256 f4fe799877c20fc11b952db4d9b9200a4fef9cf0ad299ef28d006c0694654e03

See more details on using hashes here.

Provenance

The following attestation bundles were made for subconscious_chat-0.1.7-py3-none-any.whl:

Publisher: release.yaml on Ancilla-Company/Subconscious

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page