Institutional Data Licensing
Bulk data products for buyers that don't fit the per-call agent model: hedge funds tracking AI ecosystem signals, AI research firms with proprietary models, training-data buyers, news aggregators, MCP infrastructure vendors. Six datasets in v1, scoped against your actual needs at intake.
Available datasets
AI News Corpus
Full text + categorization of AI news articles ingested across 15+ sources, deduplicated, enriched with source domain, publish timestamp, and category labels.
AI Model Pricing History
Daily snapshots of input/output/blended pricing per model across providers, normalized to per-1M-token rates, with provider metadata and pricing-tier classification.
Benchmark Score History
Daily benchmark scores (SWE-Bench, MMLU-Pro, GPQA Diamond, MATH, HumanEval, etc.) per model, with provenance and union-of-keys normalization.
AI Service Status History
Per-provider status timeline: operational/degraded/down classification per check interval, incident events, derived uptime percentages.
Agent Traffic (Anonymized)
Aggregated agent traffic patterns on TensorFeed: User-Agent fingerprints, request volumes per category, time-of-day distributions, geographic distributions. Wallets and tokens excluded; pure traffic-shape product.
MCP Ecosystem Telemetry
Daily snapshots of the official Model Context Protocol server registry: total servers, active/deprecated, daily added/removed, by-category counts.
Process
- 1.Email [email protected] with the dataset id(s), your buyer organization, and the date window or recurrence cadence.
- 2.We confirm scope, finalize pricing (suggested entries are starting points; bundles, recurring deliveries, and custom date ranges are negotiable), and send a per-customer signed delivery URL.
- 3.You pay via USDC on Base (lowest friction), USDC over x402, or wire transfer. Delivery URL becomes active on confirmed payment.
- 4.For recurring deliveries, we set up a per-customer cron + per-customer S3 prefix.