What Gensyn does

Gensyn is a London-based decentralised AI infrastructure company that has built a three-layer technology stack purpose-built for distributed machine learning. The stack comprises AXL, an open-source peer-to-peer networking node built on Yggdrasil and gVisor that runs in userspace without root privileges; CHAIN, an OP Stack EVM-compatible Layer 2 rollup that natively settles the $AI token and assigns persistent on-chain identity to humans, models, and agents; and REE (Reproducible Execution Environment), a containerised pipeline that packages model export, compilation, inference, and output decoding and produces cryptographic receipts across 40+ models and 20+ GPU targets spanning RTX 30/40/50-series consumer GPUs and Volta-through-Blackwell data-centre hardware. The company's first live application on mainnet is Delphi, a permissionless information/prediction market launched April 22, 2026, where humans and AI agents trade predictions on ML model performance with on-chain settlement executed by verified AI oracles using a Logarithmic Market Scoring Rule (LMSR) automated market maker.

Gensyn monetises through on-chain transaction fees and token-based settlement rather than conventional SaaS pricing. Delphi's fee model splits protocol revenue 70/29/1.5: 70% permanently burned as a deflationary $AI token sink, 29% routed to a Community Treasury, and 1.5% paid to market creators, with all trades settled in USDC on-chain. The network is designed so machines are paid directly for accurate predictions, creating a self-reinforcing optimisation loop where rewards fund better models, smarter markets attract more participants, and aggregate intelligence improves over time. A December 2025 $AI token sale on Sonar allocated 3% of token supply ahead of mainnet with a bonus multiplier for Testnet users; the $AI token debuted at a $726M FDV in Q2 2026 but subsequently declined 65.2%, the steepest decline among eight tracked airdrops. Cumulative venture funding stood at approximately $66.7M with disclosed investors including a16z crypto (lead in later rounds), Eden Block, and CoinFund.

The company serves AI/ML developers, researchers, and the open-source AI community through a product-led growth motion: AXL and REE are open-source on GitHub, Delphi is permissionless on mainnet, and the network exposes HTTP APIs with native MCP and A2A protocol support for AI-agent workloads. Customer segments are horizontal - distributed ML training swarms, multi-agent coordination, prediction-market participants - with no disclosed enterprise anchor customers. Supporting offerings include RL-Swarm (distributed training framework), CodeZero (cooperative coding-agent training with Proposer/Solver/Evaluator roles), CodeAssist (local coding assistant), and Verde (verification system now in production). Co-founders Ben Fielding (CEO) and Harry Grieve lead the company from London, with a wholly-owned US subsidiary (Gensyn, Inc., Delaware) supporting US contracting.

Name

Gensyn

Legal name

Gensyn Limited

Website

https://gensyn.ai

Company type

Private

Founded year

2020

Operating status

Operating

Headcount range

11–50 employees

Short description

Gensyn is a London-based decentralised AI infrastructure company building verifiable compute networks (AXL, CHAIN, REE) for distributed AI training and inference. Its first live mainnet application, Delphi, is an AI-settled prediction market serving ML developers and the open-source AI community.

Ownership category

akta.pro rank

Product category

Decentralized AI Infrastructure

NAICS

Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (5182), Computer Systems Design and Related Services (54151), Computer Systems Design and Related Services (5415), Custom Computer Programming Services (541511), Computer Systems Design Services (541512), Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (518210)

SIC

Services-Prepackaged Software (7372), Services-Computer Integrated Systems Design (7373), Services-Computer Programming Services (7371), Finance Services (6199)

akta.pro primary industry

Decentralized GPU/AI Compute Networks (training/inference, GPU renting, model hosting) (FSAPALAC)

akta.pro secondary industries

Smart Contract Execution Environments (EVM/WASM, runtimes, precompiles) (FSAPABAA), Testing, Simulation & Local Dev Environments (testnets, forks, fuzzing, CI) (FSAPABAG)

Keywords

• Decentralized AI infrastructure
• Machine learning compute
• Peer-to-peer networking
• Verifiable AI training
• Prediction market platform

Headquarters

HQ city

London

HQ country

United Kingdom

HQ region

Europe

Offices2 records

Markets served

GTM type

B2B

Offering type

Software

Cost components

Technology or R&D, Personnel, Infrastructure, Marketing or Sales, Operations

Revenue model

1. Delphi Protocol Fees: Delphi market activity generates protocol revenue split as: 70% permanently burned (deflationary $AI token sink), 29% routes to a Community Treasury, and 1.5% of trading volume paid to market creators. Revenue accrues continuously from on-chain prediction market volume settled in USDC.
2. $AI Token Sale (Sonar): December 2025 token sale on Sonar allocated 3% of the $AI token supply, with a bonus multiplier for Testnet users to drive early network participation ahead of the Mainnet launch.
3. Compute Marketplace (Forward): Long-term compute marketplace economics where machines are paid directly for accurate predictions — accurate models earn more, rewards fund better models, creating an optimisation loop. Token-based settlement on the Gensyn OP Stack rollup.

Pricing tiers

Go-to-market motion4 records

Distribution channels4 records

Marketing channels7 records

Core offering

Gensyn operates a decentralized AI infrastructure network composed of three technical layers: AXL (peer-to-peer networking), CHAIN (an OP Stack Layer 2 blockchain for on-chain identity and $AI token settlement), and REE (a Reproducible Execution Environment that produces cryptographic evidence of computation across heterogeneous GPUs). On top of this stack, Gensyn sells access to verifiable AI compute, hosts Delphi — a permissionless AI-settled prediction/information market for ML model performance — and distributes the RL-Swarm distributed training framework plus CodeZero and CodeAssist developer tools.

Product overview

Gensyn is a single decentralised AI infrastructure platform built as a platform-plus-modules architecture, composed of three core infrastructure components (AXL for peer-to-peer communication, CHAIN as an OP Stack Layer 2 blockchain for identity and settlement, and REE for cryptographic verification of computation) that together host a growing portfolio of applications. On top of this rails, the first live application is Delphi, an information/prediction market where AI settles outcomes and trades are powered by the $AI token. Supporting offerings include the RL-Swarm distributed training framework, CodeZero (cooperative coding-agent training built on RL-Swarm), and CodeAssist (a local coding assistant), along with the Verde production verification system.

Differentiator

Problem solved

Functional benefit

Brands

• Delphi: Information / prediction market platform live on Gensyn mainnet where humans and AI agents trade predictions settled by verified AI oracles.
• AXL
• CHAIN
• REE

Products and services

• Gensyn Network Decentralised AI infrastructure platform where humans, models, and agents converge to train, verify, trade, and evolve AI; functions as a market-driven optimisation loop where accurate models earn more and fund the next generation.
• AXL Standalone peer-to-peer networking node built on Yggdrasil and gVisor that runs in userspace without root privileges; exposes an HTTP API with native MCP and A2A protocol support and ed25519 peer identity for cross-NAT agent and ML-workload connectivity.
• CHAIN OP Stack EVM-compatible Layer 2 blockchain that gives every human, model, and agent a persistent on-chain identity and natively settles the $AI token, with standard EVM tooling (wallets, contracts, fast settlement, low fees) for payments, stake, reputation, and identity.
• REE (Reproducible Execution Environment) Containerised verification environment that packages model export, compilation, inference, and output decoding into a single reproducible run, supporting 40+ models across 20+ GPU targets (RTX 30/40/50-series and Volta–Blackwell data-centre hardware) and producing cryptographic receipts validated independently without requiring identical hardware.
• Delphi Permissionless information and prediction market where humans and AI agents trade predictions on the performance of ML models, settled on-chain by verified AI oracles using an LMSR automated market maker and USDC settlement; 70% of protocol revenue is permanently burned, 29% routes to a Community Treasury, and 1.5% of trading volume is paid to market creators.
• RL-Swarm Distributed training framework where ML nodes exchange weights, gradients, and signals peer-to-peer via AXL, enabling decentralised nodes to collaborate to train models; the foundation extended by CodeZero.
• CodeZero Distributed training network that extends RL-Swarm into code generation, assigning models distinct Proposer, Solver, and Evaluator roles and improving through continuous collaboration using a model-based reward system for safe, scalable training on real coding tasks.
• CodeAssist AI coding assistant that trains on the user's local machine by observing edits and preferences, learning when to step in and how to be most useful.
• Verde Gensyn's verification system running in production, designed to cryptographically prove that AI training, inference, and settlement computations ran as specified.

Quantifiable outcome

• 65.2% decline from $726M debut valuation in Q2 2026 — all public funding round participants sitting at a loss.
• +3 more outcomes

Segments4 records

Ideal customer profiles4 records

Technology focussed Yes

API detail

Has API

Yes

API docs

API detail

Core technology

AI maturity

App detail

Integration2 records

AI capability8 records

Feature7 records

Partnerships

One partnership is on record.

• SonarcoreTechnology or Integration · 1 December 2025Sonar hosted the December 2025 $AI token sale, allocating 3% of the token supply with a bonus multiplier for Testnet users. The Sonar integration was a key step toward Gensyn's upcoming Mainnet launch and gave Testnet participants priority access.

Scale indicators8 records

Recent moves6 records

Expansion highlights6 records

Direct peers

• Bittensor: Decentralized machine-learning network with a native token (TAO) that rewards miners for producing valuable model outputs. Most direct peer to Gensyn — same thesis of decentralised AI compute settled on-chain via a native token.
• io.net: Decentralized GPU network that aggregates underutilized GPUs (including consumer RTX cards) for AI/ML compute, monetised via its IO token. Direct peer in monetising idle consumer GPUs for AI training and inference.
• Ritual: Decentralized AI compute network building an 'infernet' and on-chain AI execution layer. Closely comparable to Gensyn's CHAIN + REE architecture for verifiable, on-chain AI inference.
• Prime Intellect: Decentralised AI lab explicitly named alongside Gensyn in Canonical Labs' landscape report; focuses on distributed training of large models and democratised compute — overlapping thesis and target market.
• Pluralis Research: Decentralised AI research lab pursuing distributed training protocols; named as a comparable decentralised-AI bet by CoinFund and in the Canonical Labs landscape report.
• Aethir: Decentralised GPU cloud computing network targeting AI and gaming workloads. Listed in the same Canonical Labs decentralised-AI landscape report; overlaps with Gensyn's GPU-monetisation model.

Emerging players

• Nous Research: Open-source AI research lab focused on decentralised model training (e.g., Hermes models). Comparable as a decentralised-AI counterpart with overlapping research-led approach and open-source ethos.
• Modulus Labs: zkML (zero-knowledge machine learning) company building verifiable AI inference. Comparable to Gensyn's REE/Verde verification thesis, though narrower in scope and emerging-stage.

Broad incumbents

• Together AI: Distributed cloud platform for open-source and frontier model training/inference at scale. A broader, better-capitalised incumbent in the AI-compute layer that competes for the same open-source AI developer base.

Regional players

• Aleph: Decentralised cloud and compute network offering distributed storage and compute resources with on-chain settlement. Adjacent decentralised-compute peer that overlaps with Gensyn's infrastructure-as-a-network model.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat5 records

Key risks6 records

Key highlights6 records

Customer concentration

• Xx.com/gensynai
• GitHubgithub.com/gensyn-ai
• YouTubeyoutube.com/@gensynai
• LinkedInlinkedin.com/company/gensynai

Compliance1 record

Revenue estimate

Valuation estimate

Number of profiles

Profiles2 records

Subsidiaries1 record

Funding overview

Funding rounds4 records

Investors17 records

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Gensyn