Cache API Responses to Avoid Redundant Calls

If your workflow calls an API whose response does not change frequently, cache the response and serve from cache on subsequent runs. Use a Google Sheet, database table, or even a static file as your cache layer. Include a TTL (time-to-live) so stale data refreshes automatically.

Real-world example: A workflow converts prices from USD to EUR. Exchange rates change daily, but the workflow runs every time a product is viewed -- hundreds of times per day. Cache the exchange rate with a 24-hour TTL.

// Code node: "Check Exchange Rate Cache"
const cacheKey = 'USD_EUR';
const cacheTTLHours = 24;

// $json comes from a Google Sheets or DB lookup for the cache entry
const cacheEntry = $input.first().json;

if (cacheEntry && cacheEntry.rate && cacheEntry.cached_at) {
  const cachedAt = new Date(cacheEntry.cached_at);
  const ageHours = (Date.now() - cachedAt.getTime()) / (1000 * 60 * 60);

  if (ageHours < cacheTTLHours) {
    // Cache hit -- return cached rate
    return [{
      json: {
        rate: cacheEntry.rate,
        source: 'cache',
        cached_at: cacheEntry.cached_at,
        expires_in_hours: Math.round(cacheTTLHours - ageHours)
      }
    }];
  }
}

// Cache miss or expired -- flag for refresh
return [{
  json: {
    rate: null,
    source: 'miss',
    needs_refresh: true
  }
}];
```text
Use an IF node to branch: cache hit skips the API call, cache miss calls the exchange rate API and then updates the cache:

```json
// HTTP Request node: "Fetch Fresh Rate" (only on cache miss)
{
  "method": "GET",
  "url": "https://api.exchangerate-api.com/v4/latest/USD",
  "options": { "timeout": 5000 }
}
```text
```javascript
// Code node: "Update Cache" (after fetching fresh rate)
const freshRate = $input.first().json.rates.EUR;

return [{
  json: {
    key: 'USD_EUR',
    rate: freshRate,
    cached_at: new Date().toISOString()
  }
}];
// Then write this to your Google Sheet or database
```text
```yaml
Cost savings:
  Without cache: 500 API calls/day x $0.001 = $0.50/day = $15/month
  With cache:    1 API call/day   x $0.001 = $0.001/day = $0.03/month
  Savings: 99.8%
```text
This pattern works for any data that changes infrequently: geocoding results, company info lookups, configuration fetches, or any reference data.

**Related:** [Use Structured Output (JSON Mode) for Parseable Responses](../ai-and-llm-integration/01-use-structured-output-json-mode-for-parseable-responses.md) | [Configure Payload Size and Binary Data Mode for Large Files](../performance-and-large-files/01-configure-payload-size-and-binary-data-mode-for-large-files.md)