Company: Naptha.AI
CEO & CSO: Mark Schmidt, Richard Blythman
Website: http://www.naptha.ai
About Naptha.AI
Naptha.AI is an open-source AI infrastructure platform for autonomous agents, designed to reshape how artificial intelligence is built, shared, and scaled. Founded in 2024 by Richard Blythman and Mark Schmidt, Naptha is building a decentralised internet of agents, where billions of intelligent agents can interact with each other across networks without the gatekeeping of centralised cloud providers or Big Tech intermediaries.
Our vision is to shift from the monolithic AI architectures of today, which consist of giant., closed source models in massive energy-intensive data centers, to a decentralised network of agents that are modular, secure, and efficient. In the simplest sense, Naptha is creating a system where developers can deploy intelligent agents that reason independently, collaborate in swarms, and respect user-owned data as a feature of their design.
Since launching, Naptha has raised $6 million in pre-seed funding, with more to come in 2025, and built an international team spanning distributed systems, AI safety, and multi-agent research. Most recently, we’ve launched AutoMCP, a tool that lets developers convert any AI agent into a Model Context Protocol (MCP) server with just a few lines of code. This advancement, coupled with our pre-existing multiple agents across different servers, which have been around for almost a year more than MCP itself, positionsour agents to seamlessly integrate into live ecosystems, share real-time context, and act autonomously across networks. We launched v1 of our product in December, and are expanding upon that by building products on top of the platform, of which AutoMCP is the first.
We started Naptha because we saw the risks and limits of a centralised AI future. AI shouldn’t be a black box controlled by a handful of tech giants. It should be something open, auditable, and usable by communities, researchers, and companies of all sizes. Naptha helps make that possible by allowing developers to build and test agents locally, then scale them globally across networks they trust.
A major focus for us has been ensuring decentralised AI systems can match or exceed the performance and coordination of cloud-native platforms. We’re tackling this with research into agent collaboration protocols, memory-efficient reasoning modules, and resilient communication layers that work even when connectivity is limited. This approach is centred on our commitment to rethinking how AI infrastructure can work, rather than simply mimic the centralised/closed-source models of the established giants.
The impact is already visible: early adopters are using Naptha to explore persistent multi-agent simulation and AI networks where agents adapt continuously to local environments. We’ve seen strong interest from the developer and academic communities, all looking for an alternative to the current AI status quo.
The trajectory of AI development is currently rushing towards closed-source, centralised models that are opaque and out of user control. Naptha goes in the opposite direction, offering a platform that is developer-led and decentralised in nature. We truly believe that this is the safer, more inclusive, and more scalable path forward.
