Glossary

What these metrics mean and why they matter

Core Indices

Compute CPI $CPI

The inflation rate for AI work. Tracks the cost of a standardized basket of AI workloads against a January 2025 baseline. A value of 87 means AI work is 13% cheaper than baseline.

Like the consumer price index tracks groceries, we track prompts. If $CPI is falling, your AI budget goes further.
Reasoning Tier $JUDGE

Cost per million tokens for models optimized for complex reasoning, chain-of-thought, and judgment tasks. Includes models like o1, o3-mini, DeepSeek-R1.

The "cognitive premium" tier. Users pay more here because the tasks require capabilities budget models can't match.
Frontier Tier $FRONT

Cost per million tokens for the most capable general-purpose models. Includes Claude Sonnet, GPT-4o, Gemini Pro.

Budget Tier $BULK

Cost per million tokens for high-throughput, cost-optimized models. Includes GPT-4o-mini, Gemini Flash, Claude Haiku.

The commodity tier. Prices here fall fastest because competition is fiercest.
Long Context Tier $LCTX

Cost per million tokens for models optimized for large context windows (100K+ tokens). Used for document processing, RAG, and analysis of large codebases.

Market Intelligence

Quality-Adjusted Price QAP

How much you pay per unit of model quality. Lower QAP means better value - more capability for your dollar.

A model with QAP of 0.5 gives you twice the value of one with QAP of 1.0. Flash models typically have the lowest QAP.
Cognitive Arbitrage

Whether a model is over or underpriced relative to its capabilities compared to market average. Values above 1.2 suggest the model is underpriced (buy signal). Below 0.8 suggests overpriced.

High arbitrage scores reveal market inefficiencies - models that deliver more capability than their price suggests. These are the "value stocks" of AI.
Arena ELO

Quality rating from the Chatbot Arena, where humans compare model outputs head-to-head. Higher ELO means the model wins more comparisons. Based on millions of human votes.

Market Share Velocity MSV

How fast a model is gaining or losing market share week-over-week. Positive MSV means growing adoption. Derived from OpenRouter volume data.

Spreads

Cognition Premium $COG-P

The price difference between frontier and budget tiers. Measures how much extra you pay for top-tier capability vs commodity.

A widening cognition premium suggests frontier models are holding value while budget models commoditize. A narrowing spread means budget models are catching up in capability.
Judgment Premium $JDG-P

The price difference between reasoning and frontier tiers. Measures the premium users pay for extended thinking and complex reasoning.

Build Cost Index

Startup Builder $START

Cost index for a typical startup workload: heavy on coding assistance, some RAG/context, and light routing/classification.

Agentic Team $AGENT

Cost index for autonomous agent workloads: dominated by reasoning/thinking, with tool use and structured output.

Agentic workloads are reasoning-heavy. If $AGENT is rising while $CPI falls, reasoning models are holding premium while commodity deflates.
Throughput $THRU

Cost index for high-volume processing: bulk extraction, classification, and summarization at scale.

Coming Soon

Token Flow Index $FLOW

Weekly token velocity across tiers. Measures whether overall compute demand is accelerating or decelerating.

Tier Migration Index $SWITCH

Tracks whether users are trading up to premium tiers or down to budget. "Premiumizing" means money moving up-market. "Commoditizing" means the opposite.