Use the Data

Back to Source Data

All of the data on this site is free to download, with no sign-up. To see how it was gathered, read Source Data.

May I reuse it?

Yes, you're welcome to. A formal license has not been chosen yet. If you use the data, please say where it came from and link back, so people can check the source.

Data files

The simplest way in: the same JSON files this site reads. They need no key.

• https://medicaidrewards.com/data/states/MI.json: one file per state (swap in any two-letter state code), with its plans, programs, counties, rewards, sources, and changes.
• https://medicaidrewards.com/data/directory.json: all 50 states and DC, with each state's Medicaid program name and how its plans are organized.
• https://medicaidrewards.com/data/rewards.json: the list of reward types, so the same reward can be matched across plans.
• reward_categories.json, qualifier_labels.json and federal_benefits.json in the same folder: reward categories, labels for who a reward is for, and the benefits every Medicaid member gets.

Each state file starts with exported_at, the date the file was made.

API

For custom queries, use the read-only REST API.

• Address: https://mqbnsjkeucvumdoqdeww.supabase.co/rest/v1/
• Every request needs the public key in an apikey header. The key is the same for everyone and is safe to share: eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpc3MiOiJzdXBhYmFzZSIsInJlZiI6Im1xYm5zamtldWN2dW1kb3FkZXd3Iiwicm9sZSI6ImFub24iLCJpYXQiOjE3ODAxODUyNjgsImV4cCI6MjA5NTc2MTI2OH0.s3rp33fqEmaBgO81gVDVel14g5fBjvPgsguUiWBvi8E
• Main endpoint: reward_comparison, one row per reward, with its plan, state, category, amounts, and who qualifies.
• Lists: api_states, api_plans, api_categories, api_universal_rewards.
• Status: rpc/api_health.

The API runs on a free database that can go to sleep when idle. If it is asleep, use the data files.

What to expect

• No promise that the files or the API will always be up, or that fields won't change.
• No support.
• Found a mistake in the data? See Corrections.

Endpoint reference

Generated from the project's API documentation. Add the two headers from the API section above to every request. Query syntax is PostgREST's: ?state_id=eq.OH, ?select=plan_name,reward_name, ?limit=100&offset=0.

GET /reward_comparison — Full reward data

The primary endpoint. Returns one row per reward, fully denormalized with plan, category, state, limits, and eligibility data. This is the public API contract — column names will not be renamed or removed.

reward_limits element

activities element

Only present on rewards with structured per-activity breakdowns. Most rewards will have activities: null.

reward_eligibility element

GET /api_states — State directory

One row per state. Use this to build a state picker or discover which states have data.

Disclosure tiers

Computed from live data — updates automatically as coverage improves.

GET /api_categories — Benefit categories

One row per category, ordered by display position.

GET /api_plans — Insurer directory

One row per insurer per state, with reward counts and contact info. Column names use plan_* for backward compatibility (v9 entities are "insurers" internally).

GET /api_universal_rewards — Federal benefits

Federal mandates that apply to all Medicaid members regardless of plan (NEMT, Free Cell Phone). Sourced from the federal_benefits table.

[
  {
    "id": "...",
    "reward_name": "Non-Emergency Medical Transportation (NEMT)",
    "category": "Transportation",
    "description": "Free rides to and from medical appointments...",
    "authority": "42 CFR § 431.53",
    "is_active": true
  }
]

GET /rpc/api_health — Health check and data stats

Returns API status, data freshness, and endpoint list. No parameters needed; GET and POST both work.

{
  "status": "ok",
  "api_version": "2.1",
  "data": {
    "total_rewards": <live count>,
    "total_reward_definitions": <live count>,
    "total_universal_rewards": <live count>,
    "total_plans": <live count>,
    "total_states": <live count>,
    "total_sources": <live count>,
    "total_extractions": <live count>,
    "categories": <live count>,
    "newest_update": "<timestamp>",
    "oldest_record": "<timestamp>"
  },
  "endpoints": [
    "reward_comparison",
    "api_states",
    "api_categories",
    "api_plans",
    "api_universal_rewards",
    "rpc/api_health"
  ]
}

api_health() runs as SECURITY INVOKER, so every count reflects only what the anon role can read under RLS — it is not a privileged snapshot of the full table.