curiso  by metaspartan

AI canvas for deep thought exploration and AI service integration

Created 1 year ago
250 stars

Top 100.0% on SourcePulse

GitHubView on GitHub
Project Summary

Curiso is an infinite canvas application designed for deep thought exploration and AI-assisted idea generation. It empowers advanced users by providing a visual, node-based interface to connect ideas and seamlessly integrate with a wide array of AI services, both cloud-based and local. The platform aims to facilitate richer, more accurate AI interactions by allowing users to guide conversations and organize thoughts on an unlimited workspace.

How It Works

Curiso employs an infinite canvas architecture where users create and connect nodes representing thoughts or conversation threads. This node-based system allows for intuitive visual organization and exploration of complex ideas. It achieves its power through deep integration with numerous AI providers, including OpenAI, Anthropic, Google, Groq, and local inference solutions like Ollama and LM Studio. This flexibility enables users to leverage preferred AI models and configurations, enhancing the depth and accuracy of AI-driven insights.

Quick Start & Requirements

Curiso is available as a downloadable application for Windows, macOS, and Linux. For developers, prerequisites include the Bun runtime (v1.2.0+), Rust (v1.86.0+), and the Tauri CLI. Installation involves cloning the repository, installing dependencies with bun install, and running the development build with sudo bun run desktop. Users will also need API keys for cloud AI services or locally installed AI inference servers.

Highlighted Details

  • Supports Windows, macOS, and Linux operating systems.
  • Features an infinite canvas with a node-based conversation system for visual idea organization.
  • Integrates with multiple default AI providers (OpenAI, Anthropic, Google, Groq, etc.) and local inference providers (Ollama, Exo, Jan.ai, LM Studio, vLLM).
  • Includes Retrieval Augmented Generation (RAG) support with local embedding models and a local VectorDB (IndexedDB).
  • Supports vision models, allowing image uploads for AI analysis.
  • Provides secure, local encrypted storage for API keys.

Maintenance & Community

The project is created by Carsen Klock (@metaspartan, @carsenklock). Community engagement and contributions are welcomed via GitHub issues and pull requests. Social channels include X, Telegram, and Discord.

Licensing & Compatibility

This project is licensed under the MIT License, which is permissive for commercial use and integration into closed-source projects.

Limitations & Caveats

A known issue exists on Windows where CORS errors can occur when connecting to a local Ollama instance, with a provided workaround. The associated CUR token is described as an experimental cryptocurrency with inherent risks, no guarantees of future utility or value, and is provided "AS IS" without warranty.

Health Check
Last Commit

1 year ago

Responsiveness

Inactive

Pull Requests (30d)
0
Issues (30d)
1
Star History
0 stars in the last 30 days

Explore Similar Projects

Feedback? Help us improve.