What is Shinkai?

Build powerful AI agents in minutes—no coding experience needed. Native crypto support included.

Define their tasks, schedule actions, and even have Shinkai’s AI write custom code for you.

Check out the video below to see how easy it is to create and test an agent in action.

​

Shinkai’s Github Repositories

​

Key Features

The key features of Shinkai are the following ones:

1. Agent Creation
2. Code Generation
3. Chat and Tool Execution
4. MCP Creation
5. Flexible AI Models
6. Decentralized Network
7. Shinkai AI Store

​

Agent Creation

Creating an agent in Shinkai is like building your own specialized AI assistant, but with superpowers. You start by selecting a base AI model - whether that’s GPT-4, Claude, Gemini, a local model through Ollama, or even Shinkai’s free trial AI that lets you experiment without any upfront costs. You then customize your agent with the specific tools and capabilities it needs. The real magic happens in how versatile these agents can be: you can schedule them to run automated tasks at any interval, perfect for monitoring Twitter feeds, checking email inboxes, scanning for trading opportunities, or any other recurring task that needs consistent attention.

What sets Shinkai’s agent creation apart is its interactive development environment. As you build your agent, you can test it in real-time using the sidechat feature, allowing you to refine its behavior and ensure it works exactly as intended before deployment. Need a custom tool that doesn’t exist yet? Simply describe what you need to the AI, and it can help generate the necessary code, which you can then add to your agent’s toolkit. This extensibility means your agents can grow more capable over time. You can also create agents with specific memory configurations, custom prompts, and even chain them together for complex workflows - for example, having one agent process data while another generates reports from those results. The best part is that once configured, these agents integrate seamlessly with Shinkai’s chat interface, making all these powerful capabilities accessible through natural conversation.

There are different ready-to-use tools for different purposes, for example:

• Information retrieval: Like summarizing web content, answering questions with citations, and searching the web (DuckDuckGo, Perplexity).
• Data processing: Tools for converting data formats (JSON to Markdown), running algorithms, and evaluating mathematical expressions.
• Crypto & Blockchain: Tools for interacting with platforms like Jupiter Solana, and EVM based chains (getting balances, sending transactions, etc.).
• Other functionalities: Tools for getting weather information, downloading webpages, and summarizing YouTube videos.

​

Code Generation

Shinkai transforms the way you create tools and agents by making code generation intuitive and conversational. Instead of starting from scratch, you can simply describe what you want to build - whether it’s a tool to analyze market data, a Twitter bot, or a custom file processor - and let Shinkai’s AI help you generate the necessary code. The system understands both high-level concepts and specific technical requirements, so you can start with a broad idea and refine it through natural dialogue until you have exactly what you need.

What makes this approach powerful is its flexibility and immediate feedback loop. As you discuss your requirements, you can see the code being generated in real-time, test it instantly in the integrated development environment, and refine it through further conversation. Need to add error handling? Want to support a new data format? Just ask. The AI can help modify the code while maintaining best practices and proper documentation. This capability extends beyond just writing code - you can generate entire tools that integrate with Shinkai’s ecosystem, create specialized agents with custom functionalities, or build reusable components that other team members can leverage. And since Shinkai provides a free trial AI, you can start experimenting with code generation right away without needing to configure external AI providers. Whether you’re an experienced developer looking to accelerate your workflow or someone new to coding who needs help building tools, Shinkai’s code generation capabilities make the process accessible and productive.

​

Chat and Tool Execution

Shinkai offers a modern, real-time chat experience designed to make complex work discussions feel natural and productive. At its core, you get instant message delivery and streaming responses, so you can see AI responses and team messages as they’re being typed, creating a fluid and interactive experience.

What sets Shinkai apart is its powerful integration of tools and AI capabilities directly in chat. You can seamlessly share and discuss files, code, and documents while working with various AI models like GPT-4, Gemini, Claude, or local models through Ollama. The chat interface serves as a command center where you can invoke specialized tools, run code, analyze data, or even manage spreadsheets - all through natural conversation. AI agents can be customized with specific tools and capabilities, allowing them to help with tasks ranging from code analysis to data processing. Whether you’re coding, problem-solving, or collaborating with AI assistants, Shinkai’s chat system keeps everything organized and accessible in one place, with conversations neatly organized into searchable inboxes that you can easily export when needed.

​

MCP Creation

Shinkai makes it remarkably simple to turn your tools and agents into Managed Cloud Procedures (MCPs). Once you’ve created a useful tool or configured a powerful agent, you can expose it as an MCP with just a few clicks, making it available as a cloud service that can be accessed through simple API calls. This means any functionality you build - whether it’s a data processing pipeline, a market analysis tool, or a specialized AI assistant - can be easily shared and integrated into other applications.

The beauty of MCPs in Shinkai lies in their simplicity: there’s no need to worry about complex deployment processes or infrastructure management. Your locally tested tools and agents can be transformed into production-ready cloud services while maintaining all their capabilities and integrations. Whether you’re building internal tools for your team or creating services for external users, MCPs provide a straightforward path from development to deployment, making your Shinkai-built solutions accessible wherever they’re needed.

​

Flexible AI Models

Shinkai takes the complexity out of AI model selection and setup by offering a flexible, no-lock-in approach to AI integration. Right from the start, you can begin working with our free trial AI model - no configuration needed, just instant access to get you up and running. As your needs evolve, you have the freedom to connect to powerful cloud-based models like GPT-4, Claude, and Gemini using your own API keys, giving you full control over your AI interactions and costs.

For those who prefer to keep things local or need offline capabilities, Shinkai seamlessly integrates with Ollama, making it incredibly easy to run AI models on your own hardware. With just a few clicks, you can install and run models like Llama, Mistral, or CodeLlama locally - Shinkai handles all the complex setup behind the scenes through Ollama. This hybrid approach means you can mix and match different AI models based on your specific needs: use cloud models for critical tasks that need maximum performance, switch to local models for privacy-sensitive work or cost optimization, or start with our free trial AI to experiment with features. The best part is that regardless of which model you choose, they all integrate naturally with Shinkai’s tools, agents, and chat interface, providing a consistent experience across your workflow.

​

Decentralized Network

Shinkai is built on a decentralized network that enables secure communication between AI agents through unique digital identities. Each agent can have its own decentralized identity, allowing it to interact with other agents across the network - whether they’re running locally or on remote machines. This identity system also handles payments between agents, making it possible to create automated workflows where agents can request and pay for services from each other.

The practical benefit is that you can build systems where multiple agents work together independently. For instance, you could have one agent that monitors data sources, another that analyzes the findings, and a third that takes action based on the analysis - with all the communication and potential payments happening automatically through their secure identities. While we’re still in the early stages of exploring all the possibilities this network architecture offers, it provides a solid foundation for creating more sophisticated and autonomous AI systems as the technology evolves.

​

Shinkai AI Store

The Shinkai AI Store makes sharing and discovering AI tools and agents as straightforward as possible. Using your decentralized identity, you can publish your custom-built tools and agents to the store, allowing others in the Shinkai community to benefit from your work. Whether you’ve built a data analysis tool, a trading agent, or any other useful automation, you can make it available to others with just a few clicks.

Finding and installing new capabilities is just as simple - browse through the available tools and agents, and install them instantly through deeplinks, similar to how you’d install a mobile app. See something interesting shared on Twitter? Just click the Shinkai deeplink and it installs directly into your workspace. While our store is still growing and the community is young, we’re building it to be a place where developers can easily share their innovations and users can quickly find solutions to their needs. Everything is backed by decentralized identities to ensure security and proper attribution, and we’re excited to see how the ecosystem develops as more creators join and share their work.

​

What’s Shinkai Architecture?