Iβve put together a list of AI Agent tools & frameworks.
Let me know if you know more.
Thread: π§΅
1/ PydanticAI is a Python agent framework combining type safety with FastAPI-like features. Its model-agnostic design lets you easily switch between different LLMs.
π Python
2/ CrewAI is one of the most popular AI tools, well known for multi-agent collaboration.
π Python
3/ n8n is a powerful visual workflow builder. It has over hundreds of integrations to enhance your AI agent nodes visually.
4/ Microsoft's AutoGen is another open-source programming framework that helps create multi-agent AI apps.
π Python
5/ Developed by Huggingface, Smolagents helps make powerful agents with simple tools, supporting various environments and language models.
It's goal is to create agents with only a few lines of code.
π Python
6/ Phidata helps build smart agents with memory, tools, and a user-friendly interface. They have a really nice website and well structured documentation.
π Python
7/ AgentOps helps developers create, test, and manage AI agents easily and securely.
It doesn't have a standalone user interface (UI), but it incorporates interfaces for monitoring, deployment, configuration and troubleshooting.
π Python
8/ AI SDK by vercel offers the fundamental components to build AI Agents. It integrates well with the popular ai-sdk
π΅ TypeScript
9/ Rig is a tool built on Rust that helps you build powerful AI Agents.
It's very beginner-friendly.
You can use it with many different AI models and easily connect it to other tools.
π€ Rust
10/ Eliza is a simple AI tool that helps create smart, independent agents that act consistently on different channels.
It has capabilities like document processing and media analysis, and it can be easily expanded with TypeScript support.
π΅ TypeScript
11/ LangGraph helps deploy smart applications using language models, offering cycles, control, and persistence.
It's built on top of LangChain, using the capabilities of LangChain to build more complex and advanced projects.
π Python
π΅ TypeScript
12/ Mem0 is a tool for AI Agents that improves memory and personalizes interactions.
It saves important details from user interactions and context like location, device type, and activity to give more personalized assistance.
π Python
π΅ TypeScript
π‘ JavaScript
If you enjoyed this thread, make sure to check out my newsletter for more tips and insights! instructa.ai/en
PraisonAI is a powerful Multi-Agent Framework that combines PraisonAI Agents, AutoGen, and CrewAI into a low-code solution for building autonomous AI systems
π Python
AgentScript
A TypeScript framework for building reliable AI agents, using a custom AST runtime for code-driven planning and execution.
Benefits
- Deterministic task execution
- fine-grained control (stop/start)
- Native state persistence
- Seamless human-in-the-loop workflows.
- enhanced observability.
π΅ Typescript
β’ β’ β’
Missing some Tweet in this thread? You can try to
force a refresh
I asked Gemini Advanced Pro (20$/m): "what's the best way to organise AI agents?"
See the results below.
𧡠Thread
Unlike GPT-o1/pro:
+ It presents you with a plan before taking action.
+ Plus, you can adjust the plan before it runs.
- It takes a long time to generate the results. GPT is faster here. (I guess because of the web search)
A big advantage is that Gemini Deep Research can search the web. That means we get realtime results.
If you want to do this with GPT, youβd first need to perform a web search and manually feed the results into reasoning models.
But GPT gives you much fewer web results compared to Gemini!
How to get reliable coding results with AI. Every time.
Examples for bolt, cursor, windsurf and vscode inside.
Thread: π§΅
1/ Why?
A rules file is like a guidebook for your AI coding helper. It tells the AI how to write code for your project, including what tools you're using and how everything is organized. This helps the AI create better and more accurate code.
2/ Think of ChatGPT
Well its the the same as you form prompts in ChatGPT. You don't just say
"Rewrite this text".
You form a prompt like:
"""
you are proofreading and writing expert.
RULES:
- Do not use fluff words or High-level grammar and vocabulary.
- Don't write overly long sentences to keep the text easy to read.
- Use first-person perspective and a friendly, conversational tone
1/ Supabase fills the "Backend" gap for many, as AI generators already create strong frontends.
Itβs made for developers:
- Ton of youtube tutorials
- Solid documentation (but not always 100% up to date)
- Loads of integrations
2/ Quickstart
- Sign up on supabase .com
- Create a new project to get your database link and keys.
- Use the built-in Supabase Studio to design your database.
- Add Supabase to your app to manage data and users.
- Host your app on platforms like Vercel or Netlify.