Data Sources

Attribution

Anera Markets is built on data published by the companies and communities of the AI economy. This page lists every source we collect from, what we take from each, and where it appears on the site.

Sources

61

Categories

7

6 sources

Token usage and demand

Tokens processed, requests and spend, broken down by model, token factory and application. These are the volume measurements behind every Index.

Used in Indices, Demand, Token Factories, Models, Apps, Model Families

SourceWhat we collectCollected from
Chuteschutes.aiPublic chute catalog, daily spend per chute and daily LLM token statisticsAPI
OpenCodeopencode.aiModel usage statistics and the model catalogWebsite
OpenRouteropenrouter.aiDaily tokens by model and by provider, top apps per model, app token usage, endpoint request statisticsWebsite
Replicatereplicate.comPublic model catalog, collections and run counts, from which daily spend is estimatedAPI
Requestyrequesty.aiModel usage rankings and the model data catalogWebsite
Vercel AI Gatewayvercel.comModel leaderboard ranks and token shareWebsite

17 sources

Token prices

Published list prices per million input, output and cached tokens, per model and per venue.

Used in Indices, Token Price Dispersion, Token Factories, Models, Silicon Spread™

SourceWhat we collectCollected from
Anthropicdocs.claude.comAPI token pricesPricing page
Azure AI Foundryazure.microsoft.comModel token prices from the Azure Retail Prices APIAPI
Basetenbaseten.coModel API catalog and token pricesAPI
Cerebrascerebras.aiModel catalog and token pricesAPI
Cloudflare Workers AIdevelopers.cloudflare.comToken pricesPricing page
DeepInfradeepinfra.comModel catalog and token pricesAPI
falfal.aiModel catalog and pricesAPI
Fireworks AIfireworks.aiServerless token pricesPricing page
Groqgroq.comToken prices and prompt-caching pricesPricing page
Hugging Face Inference Providershuggingface.coRouter model catalog and per-provider token pricesAPI
Inference.netinference.netModel catalog and token pricesAPI
Novita AInovita.aiLLM catalog and token pricesAPI
OpenAIopenai.comAPI token pricesPricing page
OpenRouteropenrouter.aiModel and endpoint prices, and effective (discount-adjusted) pricingAPI
SambaNovacloud.sambanova.aiModel catalog and token pricesAPI
Together AItogether.aiModel catalog and token pricesAPI
Vercel AI Gatewayvercel.comModel catalog and token pricesAPI

6 sources

Media prices

Prices for non-text generation — images, video, speech and transcription — in each venue's own unit.

Used in Media Prices, Media Price Dispersion

SourceWhat we collectCollected from
Artificial Analysisartificialanalysis.aiMedia model catalog and quality rankingsAPI
ElevenLabselevenlabs.ioSpeech and audio API pricesPricing page
Google Cloudcloud.google.comVertex AI media generation, Text-to-Speech and Speech-to-Text pricesPricing page
Replicatereplicate.comPer-model media pricesWebsite
Runwaydocs.dev.runwayml.comVideo API pricesPricing page
Stability AIplatform.stability.aiImage and video API pricesPricing page

32 sources

GPU compute prices and supply

On-demand, spot and reserved prices for GPU instances, and the capacity each venue lists as available.

Used in Compute Supply, GPU Price Dispersion, GPUs, Silicon Spread™

What we collect GPU instance prices and listed availability, by SKU, region and term

6 sources

Model metadata, benchmarks and performance

What each model is — its creator, size, architecture and modality — how it scores, and how its endpoints perform.

Used in Indices, Models, Model Families, Complexity Index, Token Factory SLAs, GPUs

SourceWhat we collectCollected from
Artificial Analysisartificialanalysis.aiModel intelligence scores, creator attribution and the number of providers serving each modelAPI
Hugging Face Hubhuggingface.coModel architecture, parameter counts and Inference Provider coverageAPI
OpenRouteropenrouter.aiEndpoint uptime, latency, throughput and benchmark scoresWebsite
Requestyrequesty.aiLatency leaderboardWebsite
SemiAnalysis InferenceXinferencex.semianalysis.comMeasured inference throughput and latency per GPU and modelAPI
Vercel AI Gatewayvercel.comDeepSecBench leaderboard resultsWebsite

1 sources

Prediction markets

Prices, open interest and daily volume of AI-related prediction markets.

Used in Prediction Markets

SourceWhat we collectCollected from
Polymarketpolymarket.comMarket prices, open interest and daily volumeAPI

4 sources

Reference data

Specifications and editorial material Anera maintains by hand from published sources.

Used in GPUs, GPU Benchmarks, Compute Supply

SourceWhat we collectCollected from
AMDamd.comGPU datasheet figures: compute, memory and configurationDatasheet
Lenovo Presslenovopress.lenovo.comGPU and server product guidesDatasheet
NVIDIAnvidia.comGPU and system datasheet figures: compute, memory and configurationDatasheet
PNYpny.comGPU datasheet figuresDatasheet

Software

Software

Third-party software whose licence asks to be credited in the product.

SourceUsed forCollected from
TradingView Lightweight Charts™tradingview.comPrice and time-series charts. Copyright © 2025 TradingView, Inc.Website

About this data

Prices, usage and performance figures are collected from each source’s public API or published pages, then normalised, aggregated and in places estimated by Anera. Figures shown on Anera Markets are Anera’s measurements and are not endorsed by the sources listed here. All product names, logos and trademarks are the property of their respective owners.

Spotted a missing or outdated source, or want your data credited differently? Contact us at [email protected].