Compassion Benchmark
Public data endpoints

Machine-readable benchmark data

Compassion Benchmark publishes score data for all 1,156 entities as structured JSON files served directly from this domain. They are free to access; if you use them in research or reporting, please cite the source with attribution. See /data-licenses for formal usage terms.

Independence note: Entities never pay for inclusion, score changes, or suppression of findings. Published scores reflect the benchmark's independent methodology only.

Quick links

Usage termsMethodology

Per-index aggregate files

One JSON file per index containing all entities, ranks, scores, and band classifications for that index.

Base URL pattern: https://compassionbenchmark.com/data/indexes/{slug}.json

Fortune 500

Browse index →

447 large U.S. companies scored across 8 dimensions.

/data/indexes/fortune-500.json

World Countries

Browse index →

193 sovereign and territorial entities — global coverage.

/data/indexes/countries.json

U.S. States

Browse index →

21 U.S. states scored in the current publication cycle.

/data/indexes/us-states.json

AI Labs

Browse index →

50 leading AI laboratories with HQ and sector.

/data/indexes/ai-labs.json

Robotics Labs

Browse index →

50 humanoid robotics labs with category and country.

/data/indexes/robotics-labs.json

U.S. Cities

Browse index →

144 U.S. cities with region classification.

/data/indexes/us-cities.json

Global Cities

Browse index →

250 global cities with country and region.

/data/indexes/global-cities.json

Aggregate file schema

Each per-index file is a JSON object with the following top-level fields:

  • generatedAt — ISO 8601 timestamp of last build
  • index — index slug (e.g. "fortune-500")
  • entities[] — array of entity objects, each with: slug, name, rank, composite (0–100 score), band (Critical / Developing / Functional / Established / Exemplary), and dimension scores (AWR, RES, POL, SYS, PAR, LRN, COM, INT)

Per-entity score files

A separate JSON file for each of the 1,156 scored entities, keyed by slug.

Base URL pattern: https://compassionbenchmark.com/data/scores/{slug}.json

There are approximately 1,156 entity files. Slugs match the URL segment on each entity's benchmark page — for example, the slug for Microsoft is microsoft and its file is at /data/scores/microsoft.json.

Entity score file schema

Each per-entity file is a JSON object used to power the live scoring badge and entity-page data. Fields include:

  • slug, name, kind — entity identifier and type (country, company, ai-lab, etc.)
  • composite — overall score (0–100), weighted mean of the 8 dimensions
  • band — band name (Critical / Developing / Functional / Established / Exemplary)
  • rank, indexTotal — position within the index and total entities in that index
  • dimensions — object with per-dimension scores (AWR, RES, POL, SYS, PAR, LRN, COM, INT)
  • generatedAt — ISO 8601 build timestamp

Entity catalog

A single file listing every scored entity across all indexes.

/data/index.json

The top-level catalog of all entities. Use this file to discover available slugs before fetching individual score files. Fields include:

  • generatedAt — build timestamp
  • totalEntities — count of all scored entities
  • indexes — array of index slugs
  • entities[] — array with slug, name, indexSlug, kind, rank, composite, band

Suggested use pattern

  1. Fetch /data/index.json to enumerate all available slugs and their index membership.
  2. Fetch individual /data/scores/{slug}.json files for the entities you need.
  3. Or fetch /data/indexes/{slug}.json to get all entities for a given index in one request.
  4. All files are static JSON served with standard cache headers — no API key required.

How to cite this data

If you use Compassion Benchmark data in research, reporting, or a product, please cite the source.

General citation format

Compassion Benchmark. "[Index or Entity Name]." compassionbenchmark.com, accessed [Month Year].

For a specific entity page, use the full URL, e.g.
compassionbenchmark.com/fortune-500/microsoft

Attribution notes

  • Data is free to access. We ask for attribution when the data appears in published work, reporting, or a product visible to others.
  • Scores reflect the benchmark's independent methodology — see /methodology for the full framework.
  • For formal data usage rights (CSV exports, institutional use, redistribution), see /data-licenses.

For press inquiries and data briefing requests, see For Press & Researchers.

Formal data licensing

Need CSV exports, structured data rights, institutional access, or multi-team licensing? See the data licenses page for usage categories and contact sales for a quote.

Data licensesContact sales