Chris Messina

SmythOS - Build, debug, and deploy AI agents in minutes

by
Top Hunter

SmythOS lets you build and deploy AI agents without manual coding. Describe your needs, and Agent Weaver creates it automatically using the best AI models and APIs. Integrate with OpenAI, Hugging Face, Amazon Bedrock, and more—no manual coding required.

Add a comment

Replies

Best
Chris Messina
Top Hunter
Hunter
🔌 Plugged in
📌
Super impressed with the ease of use and capabilities of SmythOS. Their visual builder not only makes it easy to generate cross-cloud performant agents, but the Smyth Runtime Environment is free to use; you just pay to execute agent actions making it easy to prototype and get started. Don't get locked into one platform — use the best among them for what they're good at! The most exciting unlock for me was asking Agent Weaver to create an agent for me and then watching it execute my request, including making tweaks or updates based on my conversational input.
Alexander De Ridder
@chrismessina Thank you for hunting us. The man who hunted ChatGPT and Sora also discovered SmythOS! Your legacy keeps growing. :-) SmythOS changes how the world builds & deploys AI agents! Let's go!
Another AI no-code platform. It is said that 2025 will be the dawn of AI Agents, and I hope SmythOS will become the birthplace of creating a large number of rich AI Agents. 🚀
Joseph O'sehinde
@victoria_wu Agent Weaver makes it even more exciting—you simply describe what you need, and Agent Weaver handles the rest seamlessly.
Alexander De Ridder
@victoria_wu 2025 will be amazing indeed!
Mario Braune
@victoria_wu @franklin101 yes, this does not get recognition enough, this is an absolute mindblowing feature only available at smythos IMHO
Brett Heyns
@mario_braune How much time do you think that Agent Weaver saves you a week?
Mario Braune
@brett_heyns1 i would say several hours a week at least, but mileage differs depending on the usecase : i use it mostly early in the process , for example to create the basic agent while i still flesh it out by hand later. if its a complex agent, the creation and connection of its components alone can save hours. In established, well fleshed out agents on the other hand, i use it mostly to act as a coding assistant , so i will have agent weaver look at a code component which needs tweaking for example and let it improve the code, so it replaces chatgpt in an extra tab for me here. Overall: mindblowingly good feature which does not get enough recognition
Tony Tong
🔌 Plugged in
Congrats team on the launch! The chat to AI agent UX makes sense, and the demo is siiick (shoutout to the arcade team)!
Michael
Maker
Thank you, @tonyabracadabra! Arcade is a game-changer for sure
Alexander De Ridder
@tonyabracadabra We found out about arcade when we were getting for product hunt. It really helped us showcase SmythOS, happy such product exists.
Alex Fino
Congrats in the launch! Looks great.
Michael
Maker
thank you, @alex_fino!
Alexander De Ridder
@alex_fino Thank you Alex!
Brett Heyns
What is truly revolutionary is how Agent Weaver takes your natural language prompt, "I want to create an Agent that can do XYZ", and automatically builds it for you! You can prototype and test, to your heart's content, before you are ready to deploy. And with all the really great built in integrations, you can connect with most of the popular apps, with ease. Automations, workflows, Agent Orchestration, industry-specific solutions... All are possible with the visual builder. The only limit is your imagination!
Michael
Maker
@brett_heyns1 thanks, Brett! excited to see what you build next 😎
Alexander De Ridder
@brett_heyns1 Glad you love Agent Weaver. Let's see what you cook!
Joseph O'sehinde
SmythOS is making waves! Agent Weaver is a game-changer the world needs to experience, creating an agent has never been this simple.
Michael
Maker
@franklin101 🌊 🌊 🌊 ...happy building!
EthanYu
Finally tried SmythOS after struggling with AI agent development - this is what I've been waiting for! 🚀 Love how it auto-picks the best AI model for different tasks. No more juggling between OpenAI and Hugging Face APIs manually. Built a customer support agent in 15 mins that actually works!
Alexander De Ridder
@ethanyu Thank you so much for sharing your experience! 🙏 Really happy to hear the model selection helped streamline your workflow. The 15-minute deployment time is exactly what we were aiming for - making production-ready AI agents accessible without sacrificing capability. We'd love to hear more about your specific use case and what other features would make your agent development even smoother. Your real-world feedback helps us prioritize what to build next.
Michael
Maker
thank for your support @ethanyu 😎
Brett Heyns
@ethanyu "Built a customer support agent in 15 mins that actually works!" This is The Way, and how it is supposed to be. Amazing that it's working so well for you.
Moin Uddin
Congrats on the launch everyone! Can't wait to see what the community builds with it.
Michael
Maker
@moinism thank you for the support 💪
Alexander De Ridder
@moinism Same, I love hearing the countless stories from our users. Some say it has changed their life!
Alexander De Ridder
Hey Product Hunt community! 👋 I'm excited to share SmythOS - a powerful visual builder AND runtime environment that makes building, debugging, and deploying AI agents radically easy and fast! 🚀 I began working in ML in 2008, researched sparse embeddings in 2015 which resulted in several patents, launched my first AI SaaS in 2019, and together with my team and co-founders bootstrapped SmythOS. Creating a product like SmythOS is a labor of love. We hope you join us. With SmythOS, the Wordpress moment of AI agents has arrived - what used to be complex is now easy and fast. It's time to create your own 3 human, billion dollar company. LLMs are becoming commoditized like spell checkers. AI agents is how you put them to work. Instead of building vertical agent solutions, we created the tooling so you can build the next great company. Wake up with a great idea in the shower, and 2 hours later, it is in production. SmythOS represents a watershed moment where the way we think about automation, productivity, and ROI is radically transformed. We envision a world where agents empower everyone, not just the rich, to usher in an age of soulcraft. What makes SmythOS unique: 🤖 Multi Agent & Multi Modal - Build your own Cursor or Devin-style multi agent systems with drag and drop ease. We have agent teams that help us code, analyze products, support customers, and even build other agents. SmythOS agents work with any data type (image, binary, text, code, ...) and today power our entire blog. ⚙️ Smyth Runtime Environment (SRE) - Think of it like Java Runtime Environment, but for .smyth files. Our 100% free runtime on GitHub lets you run agents on Mac, Windows, Linux. Our OS-inspired architecture handles all complex infrastructure so you focus on agent logic. At just 50MB with zero dependencies, deploy anywhere! 🧠 Agent Weaver - The world's first AI agent that builds AI agents, letting you use chat to create sophisticated flows. Powered by our reasoner-style AI model trained on synthetic data, Weaver knows how to use your existing API keys and agents (without training on them). 🔓 No Vendor Lock-in - Use any AI model, API, or data source. Build once, deploy anywhere - locally, our cloud, or your cloud. Export everything - you own your agents. 🛡️ Secure & Scalable - Hashicorp encryption, API key vault, air-gapped components, ACL, frequent audits. We have scalable proxmox clusters with multiple layers of fallback and an incredible DevOps team to take all the headaches of scale off your mind. 👥 Team Features - Full role management, spaces for teams, projects and clients. Built for seamless collaboration. 🤗 Huggingface & More - Drag & drop integration with hundreds of thousands of Huggingface AI models. Works with Zapier AI actions and hundreds of services out of the box. Plus universal API support and deep Postman integration. 🎯 Visual Building & Debugging - We have pioneered constrained alignment for complete AI control. Chain multiple AI models together. Easy debugging with comprehensive deployment logs. Debugging might be my favorite feature! 💾 Drag & Drop RAG - Bring your own vector db or use ours. Agents can remember, forget, and retrieve data from your drag & drop data lake. Simple enough for anyone. What you can build: ⚡️ API: The fastest way to create and deploy APIs, guaranteed. Perfect with v0 and Cursor. Build a SaaS/App instantly - we handle everything. 🧪 LLM: Transform your AI Agent into an OpenAI SDK-compatible LLM. Build expert mixtures, multimodal hybrids, unlimited API usage. 👨‍💼 Digital Coworker: Chat with agent teams, assign work/schedules, handle webhooks. Always ready, always learning from new AI models. 🤝 GPTs, Copilots, Chatbots: Deploy to any platform - Slack, Discord, Gems, and more. Seamless integration with your existing tools. This is just scratching the surface! SmythOS is revolutionizing how we build and ship backends, APIs, LLMs, agents, and digital coworkers. 🚀 Would love to hear your thoughts! I'm here all day to answer questions and geek out about AI agent development. Let's build something amazing together! ✨ Happy hunting! 🎯
Brett Heyns
@agent_adridder , this is an AMAZING breakdown, of all of the incredible features and benefits of SmythOS. The Smyth Runtime Environment and Agent Weaver are just mindblowing!
Menelaos Kotsollaris
💡 Bright idea
Congrats on the launch! We have an SEO integration with kwrds.ai, and we are super impressed by the capabilities of fully automated AI agents. It used to be a struggle to creat impactful content that sticks, but with this integration, keyword research and people also ask tool are managed directly from the AI agents and the kwrds.ai API. I also love the ability to plugin into several other APIs, and create custom solutions!
Alexander De Ridder
@mkotsollaris Thanks for sharing your experience with the SEO integration! 🎯 It's fantastic to hear how the combination of SmythOS agents with kwrds.ai's API has streamlined your content creation workflow. The ability to handle keyword research and "People Also Ask" data directly through AI agents really shows the power of API integrations. Your use case is a perfect example of what we envisioned - letting agents handle the complex orchestration between different services while delivering actionable results. Have you explored using our memory management system to help your agents maintain context across different content creation sessions? Could be really useful for maintaining consistent SEO strategy! Let us know if you discover any other interesting integration possibilities - we're always excited to see how users combine different APIs to create unique solutions!
Brett Heyns
@mkotsollaris Thank you for the kind words. Your integrations, are a shining example of how disparate products/tools, can now finally work well together! To now achieve great things (and reduce struggle), is super possible with incorporating amazing products such as kwrds.ai!
Samme Abdul
Agent Weaver is a game-changer! It’s incredible how effortlessly it transforms your ideas into fully functional agents—just describe what you need, and it takes care of the rest. I love how you can prototype, test, and refine until everything is perfect before deployment. Plus, the seamless integrations with so many popular apps make automating and streamlining workflows ridiculously easy. Whether you need automations, advanced workflows, or something industry-specific, the visual builder makes even the most complex ideas achievable. Honestly, the only limit is your imagination—Agent Weaver truly delivers!
Alexander De Ridder
@samme_abdul Glad to have you as a user of SmythOS!
AJAY KUMAR YADAV
You’ve earned my upvote! The ability to deploy AI agents with minimal coding is incredibly exciting — that’s a game-changer for accessibility. Regarding the name "smythOS," it does come across as just another operating system at first glance. Perhaps emphasizing the unique features or AI focus could make it stand out more. Nevertheless, I’m eager to give it a try and see how it performs. Wishing you tremendous success on this journey!
Alexander De Ridder
@ajay_kumar_yadav4 Thank you so much for the thoughtful feedback and support! You make an excellent point about the name - we've discussed internally how we could better communicate that SmythOS is specifically designed for AI agent orchestration and deployment, rather than being a general-purpose OS. We're actively exploring ways to make this clearer in our messaging. The accessibility aspect you highlighted is actually core to our mission. While there's powerful functionality under the hood (for example our ACL system and security model), we've worked hard to ensure that someone can get started deploying useful AI agents without needing to understand all the technical details. If you do try it out, we'd love to hear about your experience, especially around the ease of use and deployment process. Your feedback would be invaluable as we continue refining the platform. Thanks again for the upvote and the constructive feedback! 🙏
Gary Haymann
This visual builder and chat to agent Ai platform is the easiest tool to bring AI agent into your business ! Love it !
Alexander De Ridder
@ghaymann Thank you so much for the kind words! 🙌 We designed it specifically to make AI agent deployment accessible while keeping powerful capabilities under the hood.
Chelle Fuertes
This is it! Can't wait to see all the amazing agents everyone will build with SmythOS!
Alexander De Ridder
@chelle_fuertes Thanks for your excitement! 💫 The community's creativity with building different types of agents has been incredible to watch. We're just as excited to see what unique use cases and implementations you'll come up with! Don't hesitate to share your creations in our community forums.
Ashit from Draftly.so
Congrats on the launch, @chrismessina @agent_adridder @agentwalker4. Looks like a great tool for quick prototyping! 👍🏼️ How does it decide which AI models or APIs to use? Does it prioritize cost, performance, or something else? Just curious.
Alexander De Ridder
@chrismessina @agentwalker4 @ashitvora In the visual editor you can choose which models to use. You simply add your API key to the vault and all the models become immediately available for drag & drop. Agent Weaver (the AI agent that builds your AI agents) prioritizes API services that you have already added.
SADIK KASIMOGLU
SmythOS is exactly what the AI space needs right now! The intuitive drag-and-drop interface and the ability to create and deploy powerful AI agents without deep coding skills are game-changers for individuals and businesses alike. I have a few questions to understand the potential for leveraging SmythOS as a business solution: 1- Can I create custom AI agents for my clients and resell the service? Are there any specific tools or guidelines within SmythOS to support this model? 2- If I deploy an AI agent on AWS for my clients, how are updates handled? Will I need to re-upload new code for every update, or does SmythOS have a streamlined way to manage updates remotely? 3- Are there any best practices or recommended setups for securely connecting AI agents to client databases to ensure real-time data interaction? I'm genuinely excited about the possibilities here. Can't wait to hear your thoughts and explore how SmythOS can help scale client-specific AI solutions!
Alexander De Ridder
@esadik Thanks for your detailed questions about SmythOS! We are beyond excited to see what our community is already building. Thousands of agents get deployed every week now, and we see companies automating every process imaginable! Let me break down each point: 1. We definitely support creating and reselling custom AI agents! You can white-label both your development environment and deploy agents to your own subdomain. Many agencies are already seeing great success with this model - some even landing $100K+ projects. For API monetization, you can integrate with various Product Hunt tools or API marketplaces to help monetize your APIs, including any LLMs you create with SmythOS. 2. Updates are super streamlined! Our built-in versioning system lets you deploy updates with just a few clicks - add your description, hit deploy, and your agent is instantly updated in production. If you need any help with this, our Discord community is always ready to assist. 3. For database connectivity, agents interact with data in real-time. When a component (like an API or data component) calls a data service, it queries that source instantly - just like a human would. We support all standard security protocols including OAuth and API keys - basically anything you can do in Postman, you can do in SmythOS. This ensures secure, real-time data interaction with client databases. The platform is designed to scale with your needs while maintaining enterprise-grade security. If you have any other questions about implementing client-specific solutions, feel free to ask!
Karl Yeh
I'm curious about integrations? Obviously would love to see this in a variety of use cases from Google Drive, OneDrive, Slack, Asana, etc...
Alexander De Ridder
@karl_yeh It's all there! Google Drive, Asana, etc. We have hundreds of integrations pre-built for you. To access, in our builder, you can go to the components section, then you'll find our pre-built integrations. See image: Moreover, Agent Weaver (chat to agent) knows many APIs from around the web through its base LLM, and if you ask it to make an integration, it knows how to use the API component, OR .. you can feed it some API docs as context in the chat and let it build from there!
Anton Osipov
Congrats on the launch! It’s a powerful tool. I might not be using it exactly as intended, but I think there are opportunities to smooth out the user experience. Here are some design principles I noticed you’re applying in parts of the product but not quite consistently: Progressive Complexity It helps to show functionality only when it’s relevant. Controls could be hidden or disabled to indicate availability without creating confusion. A specific example: if the debugger isn’t available, the corresponding switch should be disabled rather than active. Otherwise, it feels like something is broken. Clarity Terms and processes should either be intuitive or come with easy-to-access explanations. Here are a few examples of where this could be improved: - Debug injection: The only way to understand how it works is by experimenting, but even then, I couldn’t figure out the purpose of the output tab or why a debug session closes after performing an injection. - The Component Integration Request form doesn’t explain what’s happening or what it’s for. - Clicking on an agent card to open preferences was surprising since there are many other buttons on the card. Restriction While customization can be powerful, too much freedom risks overcomplicating an interface. For example, in the flow diagram, I’m not sure if the ability to move inputs and outputs around or change their colors adds much value, it might just add unnecessary complexity. This one is subjective, but I find having an explicit Input and Output block in the flow diagram really helpful. It makes it easier to see what’s coming in and going out at a glance. Despite all these comments, I think Smythos has a lot of potential. Great work, and best of luck!
Alexander De Ridder
@anton_osipov Thank you for this incredibly thoughtful and detailed feedback! This kind of specific UX analysis is invaluable for improving the platform. You raise excellent points about progressive disclosure and clarity. You're absolutely right about the debugger switch - having inactive controls that appear clickable is confusing. We should be using disabled states to clearly indicate what's available in the current context. The debug injection and Component Integration Request issues highlight a broader need for better in-context documentation. We should provide: - Clear explanations of what debug injection does and how to use it effectively - Documentation for the output tab's purpose - Better communication about debug session lifecycle - Clear context for the Component Integration Request form's purpose and workflow Regarding the agent card interaction - that's a great catch about the click behavior being unexpected with multiple buttons present. We should make the interaction model more intuitive there. Your point about the flow diagram restrictions is particularly interesting. You're right that unlimited customization (like color changes and arbitrary positioning) might be adding complexity without proportional value. The suggestion about explicit Input/Output blocks for visual clarity is excellent - it would make the data flow much more immediately apparent. Would you be open to sharing more about how you're using the platform? You mentioned you might not be using it "exactly as intended" - but novel use cases often help us understand where we need to be more flexible or provide better guidance. Again, thank you for taking the time to provide such constructive feedback. We'll definitely be incorporating these insights into our UX improvements.
Anton Osipov
@agent_adridder Thank you for such a detailed response. I'm happy that it was useful. I'll be glad to talk more about SmythOS. I've just sent you a message on linkedin.
Alexander De Ridder
@anton_osipov I responded!
Brett Heyns
@anton_osipov This is incredible feedback, and hugely appreciated 🙏 From a UX perspective, what are your thoughts about key-moment tooltips? Our interface that can "learn" about user behavior, and create pop-up tooltips, at the most common bottlenecks.
Anton Osipov
@brett_heyns1 Tooltips are quite powerful and they certainly help to work with the interface, and I love that the interface can learn and adapt to user behavior. Nevertheless, these are local optimizations. Global optimization would be forming narrower flows when a user isn't distracted by functionality that is irrelevant at the current moment.
Chad Riddersen
Congrats to the SmythOS team on the PH launch :) Love the marketing use-cases (content creation agents in particular)
Michael
Maker
thank you, @chad_riddersen!
Jeremy Corman
Wow, well done, team! The simplicity, the flow, and the power behind Agent Weaver are seriously impressive. Can’t wait to see where this goes!
Rechelle Ann Fuertes
@jeremc Thank you, Jeremy! Means a lot to hear that. Agent Weaver’s been such a fun project to bring to life! We’re just getting started, and there’s so much more we can’t wait to build with it!
Alexander De Ridder
@jeremc Thanks so much! 🙌 Really appreciate the kind words about Agent Weaver. We designed it to hit that sweet spot between simplicity and power, so it's fantastic to hear that resonated with you. Stay tuned - we've got some exciting updates in the pipeline that will make agent creation even more intuitive!