OpenAgents Raises $1.3M to Build Bitcoin-Native AI with Pylon Network Paying Users in BTC

OpenAgents Secures $1.3M Pre-Seed Funding to Advance Bitcoin-Native AI Infrastructure

OpenAgents, an open-source artificial intelligence lab focused on Bitcoin-native infrastructure for machine learning, has announced its graduation from the BitcoinFi accelerator and the close of a $1.3 million pre-seed funding round. The capital will be used to expand Pylon, its distributed compute network, and accelerate development of Psionic, its Rust-based machine learning framework.

Psionic supports inference, fine-tuning, embeddings, image generation, and distributed training, positioning OpenAgents as a public, open alternative to closed AI labs.

"America needs an open-source AI lab that can compete at the frontier without recreating the closed, centralized incentives of the biggest labs," said Christopher David, founder and CEO of OpenAgents. "OpenAgents exists to build that lab in public. We are paying people directly for the compute, software, and data that make the system better."

Pylon: A Bitcoin-Native Compute Market for Gamers and Everyday PCs

Pylon serves as OpenAgents’ compute miner, running on contributors’ machines and connecting to the Nexus coordination layer. It makes local compute available to the network, with contributors paid in Bitcoin via the hosted Nexus treasury for eligible work and launch-period payouts.

This model revives the early Bitcoin mining ethos, where accessible hardware like GPUs and gaming PCs could compete. It may also introduce a new generation of gamers and AI enthusiasts to cryptocurrency through practical, earn-as-you-go participation.

Built on Open Protocols: Nostr and NIP-90 Integration

Pylon operates as a Nostr client and NIP-90-style service provider, while Nexus handles provider presence, work assignment, telemetry, payout accounting, and public statistics. Although OpenAgents currently runs a hosted Nexus, the code is open source, allowing others to deploy their own Nexus networks over time.

The near-term focus is the OpenAgents Compute Market, with live product families including inference and embeddings. Training workflows are now being introduced with explicit assignment, validation, checkpoint, and payout processes.

Public Beta Shows Rapid Growth in Compute Participation

During the public beta, Pylon activity surged from early liveness checks to real assigned work. According to a press release shared with Bitcoin Magazine:

• More than one million satoshis paid through the hosted Nexus treasury
• Over one thousand Pylon instances activated during the first wave of public participation

OpenAgents is preparing distributed training runs that will publish detailed participation data, including online contributors, assigned contributors, accepted contributors, and model-progress contributors. The goal is to scale beyond prior decentralized training demonstrations while maintaining verifiable records of assignment and acceptance.

Targeting Stranded Consumer Compute for Open AI

OpenAgents is focused on harnessing stranded consumer compute—such as Macs, gaming PCs, and older machines—whose spare capacity is typically unused. The company views this underutilized infrastructure as a critical, untapped layer for open AI development.

All training is conducted using Psionic, OpenAgents’ Rust-based machine learning framework, ensuring efficiency and compatibility with consumer hardware.

A Product Suite Above the Network

The compute network forms the foundation of OpenAgents’ broader product ecosystem, enabling decentralized, Bitcoin-incentivized AI development at scale.