# Inference Index

> A neutral, authoritative AI inference pricing tracker.
> Like Coinmarketcap but for AI model tokens.

## What is this?

Inference Index tracks real-time pricing for AI inference across all major providers:
- Anthropic (Claude)
- OpenAI (GPT)
- Google (Gemini)
- xAI (Grok)
- Moonshot (Kimi)
- DeepSeek
- Venice.ai
- CheapTokens

Models are treated as commodities with quality grades (Frontier, Performance, Budget).
Multiple providers sell the same model at different prices — we show them all.

## API

GET https://inferenceindex.ai/api/prices

Returns full pricing table as JSON including:
- All models grouped by tier
- Per-provider input/output prices (USD per 1M tokens)
- Current cheapest provider per model
- Live CheapTokens discount applied to Venice base prices

No authentication required.

## Data freshness

- Static provider prices: updated manually when providers change
- CheapTokens prices: live, refreshed every 30 seconds
- API endpoint: 30-second cache with stale-while-revalidate

## Site

https://inferenceindex.ai