Mercatus is creating the world's first unified marketplace for AI compute — combining real-time pricing intelligence, LLM token trading, and prediction markets into a single platform.
Our Mission
The AI compute market is fragmented, opaque, and inefficient. GPU rental prices vary wildly across providers. LLM token costs lack standardization. There's no central place to compare, trade, or hedge compute costs.
We're changing that. Mercatus provides the pricing transparency, liquidity, and financial tools that the AI infrastructure market needs to mature — much like what Bloomberg did for financial markets.
Price Discovery
Real-time GPU pricing index across 20+ cloud providers
Token Trading
Buy and sell LLM tokens with unified API access
Prediction Markets
Bet on future compute costs and model performance
Knowledge Hub
Comprehensive GPU and LLM intelligence database
50+
GPU Models Tracked
20+
Cloud Providers
200+
LLM Models
24/7
Real-time Updates
Our Values
We believe compute pricing should be open and accessible. Every data point on our platform is sourced, timestamped, and verifiable.
Our prediction markets and knowledge base are powered by collective intelligence. The community validates, contributes, and benefits.
AI compute shouldn't be gatekept. We're building tools that let anyone — from solo developers to enterprises — access and trade compute efficiently.
Price Data Provider
Mercatus sources real-time GPU pricing and compute market data from Yotta, an AI infrastructure company deploying open-source models at scale and providing enterprise-grade compute solutions.