Weather, crypto, and news from one endpoint.

Sprime normalizes upstream APIs into a consistent JSON shape with caching and failover built in. One key, one format.

Live JSON below, no signup required.

sprime demo
GET /api/demo/weather?location=miami

                        
Memory first caching
{} Consistent JSON shape
Clean error formats
Built for dashboards

The shape problem

I built Sprime because I kept rewriting the same API glue.

Upstream API shape example
{
  "main": { "temp": 299.82 },
  "weather": [{"id": 801}],
  "wind": { "speed": 5.14 },
  "name": "Miami",
  "cod": 200
}
Sprime response actual output
{
  "status": "success",
  "data": {
    "temp": 26.4,
    "condition": 2,
    "windspeed": 12.3,
    "location": "Miami, United States",
    "unit": "celsius"
  },
  "source": "open-meteo",
  "timestamp": "2026-02-10T..."
}

Same fields every time, so your UI stays simple.

How it works

1

Call the endpoint

Send a GET request with your API key and parameters.

2

We fetch, normalize, cache

Sprime hits the upstream provider, maps the response to a fixed schema, and caches it.

3

You get consistent JSON

Every response follows the same shape. If a provider fails, the next one takes over.

Pricing

Free

$0
forever
  • 500 requests / day
  • Weather, Crypto, News
  • 10 minute cache
  • Community support
Get an API Key

Starter

$5
per month
  • 5,000 requests / day
  • 15+ endpoints
  • Forex, Air Quality, Geocode, more
  • Email support

Pro

$15
per month
  • 20,000 requests / day
  • All endpoints
  • 2 minute cache refresh
  • Webhooks

FAQ

What providers are behind this?

Weather uses Open-Meteo (primary) and OpenWeatherMap (fallback). Crypto uses CoinGecko, CoinCap, and Binance. News uses NewsAPI and The Guardian. All provider selection is automatic.

What happens if an upstream provider is down?

Each endpoint has a failover chain. If the primary provider fails, Sprime tries the next one. If all providers fail, you get a structured error (never a stack trace or HTML page).

Is there caching?

Yes. Responses are cached in memory first, with optional Redis as a second layer. Cache TTL depends on your plan: 10 minutes (Free), 5 minutes (Starter), 2 minutes (Pro). You can bypass cache with fresh=true on weather.

How do I contact support?

Email support@sprime.us. Starter and Pro plans get priority responses.