SIGNAL
RUNS DEEP.
We build Voyons, a world model for markets. Simulate thousands of plausible futures before every decision.
Hyperextropy in Action
Market outlook from AI scenario distributions
THE SIMULATION LAYER
Duon Labs builds Voyons, a world model for financial markets. Instead of point predictions, we simulate thousands of plausible futures. Enabling planning, stress-testing, and decision-making under uncertainty.
From research vaults to institutional APIs, our simulation infrastructure powers live deployments where traditional forecasting falls short.
Research-driven. Transparent. Live.
World model for markets. Learns market dynamics and generates scenario distributions, not point forecasts. API and SDK for funds, platforms, and agents.
CoinStats
Scores tracked by 1M+ users worldwide
Hyperextropy Vault
Scenario-driven allocation on-chain
Prophetica
1000+ daily forecasts for the top crypto assets
SIMULATION INFRASTRUCTURE
EMBED THE
WORLD MODEL
A learned market simulator exposing scenario distributions, regime dynamics, and risk surfaces. For planning and evaluation, not just prediction.
-
Scenario Trees, Not Point Forecasts
Generates thousands of plausible futures per query. Explore scenario trees, quantify tail risks, evaluate decisions before execution.
-
Learned Market Dynamics
Learns directly from market data without parametric assumptions. Adapts to regime shifts, structural breaks, and tail events as they occur.
-
Production-Ready Infrastructure
REST API and Python SDK. From hedge funds to consumer apps. Hyperextropy and CoinStats run on Voyons.
Demo: Scenario simulation showing risk analysis across multiple scenarios. Try it yourself at playground.duonlabs.com
Pushing the Frontier
Open-source research powering the world model. Foundations in distribution modeling, scaling laws, and simulation infrastructure.
SICK BIDE
A novel framework for assumption-free distribution modeling that unifies continuous, categorical, and ordered distributions through GPU-accelerated probabilistic computations without relying on predefined structures.
Read MoreApogée
Charting the scaling laws of crypto market forecasting. This research quantifies predictability using an information-theoretic approach to measure how future price movements can be inferred from historical data.
Read MoreProbabilistic Market Analysis
Research combining OHLC data with on-chain metrics to create multi-scenario market analysis frameworks that account for uncertainty in financial forecasting, enabling robust risk management for crypto market participants.
Read MoreWork With Us
Partner with Duon Labs to integrate world model infrastructure into your platform, fund, or agent system.