Apify Cost Optimizer

Stop burning compute units you don't notice.

Most Apify developers lose 20-40% of their monthly spend to silent failures, timed-out runs, and over-provisioned memory. This actor connects to your account via API, analyzes every actor's run history, and tells you exactly where your money is going and how to stop losing it.

---

What it finds

• Failed run waste - actors that regularly fail after consuming CUs, with zero useful output
• Timeout waste - runs that hit the time limit, burn max CUs, and return nothing
• Over-provisioned memory - actors set to 4096MB finishing in 8 seconds (paying for RAM you don't use)
• High cost-per-run actors - outliers burning disproportionate budget
• Low success rate actors - anything below 70% success is hemorrhaging money

---

Example output

[CRITICAL] my-linkedin-scraper
  Issue : 67% of spend is from failed/timed-out runs ($12.40 wasted)
  Fix   : Add proxy rotation and retry logic. Target site likely rate-limiting.
  Save  : ~$8.68/month

[HIGH] product-price-tracker
  Issue : Memory set to 4096MB but avg run takes 11s - over-provisioned
  Fix   : Reduce memory to 2048MB - saves ~50% compute cost per run
  Save  : ~$4.20/month

[HIGH] google-maps-scraper
  Issue : 34% of runs time out - wasting $6.80
  Fix   : Split large inputs into smaller batches. Use requestQueue for pagination.
  Save  : ~$5.44/month

---

Output fields

Field	Description
actor_name	Actor display name
total_runs	Runs in the analysis period
total_cost_usd	Total spend on this actor
avg_cost_per_run	Average cost per execution
success_rate	Percentage of runs that succeeded
waste_usd	Money lost to failures and timeouts
estimated_saving_usd	Projected monthly savings if fixed
top_issue	Plain-English description of the problem
recommendation	Exact fix to apply
priority	CRITICAL / HIGH / MEDIUM / OK

Results are sorted by waste_usd descending - biggest problems first.

---

Input

Field	Default	Description
apifyToken	required	Your API token from console.apify.com/account/integrations
lookbackDays	30	Days of history to analyze (1-90)
minRunsToAnalyze	3	Skip actors with fewer runs than this
includeSucceeded	true	Include successful runs in cost analysis

Your token stays private. Used only to call the Apify API from within the actor. Never logged or stored.

---

How to get your API token

1. Go to console.apify.com/account/integrations
2. Copy your Personal API token
3. Paste it into the apifyToken input field

---

Pricing

$9/month rental. Platform usage per run is approximately 0.01-0.03 CU - essentially free on top of the rental fee.

One run typically finds enough savings to pay for itself many times over.

---

Limitations

• Memory waste detection uses allocation vs. run duration as a proxy - actual peak memory usage is not available via the Apify API
• Savings estimates are projections, not guarantees
• Requires at least minRunsToAnalyze runs per actor in the lookback period to include that actor in analysis

---

Roadmap

• Slack and Telegram alerts when a new waste pattern is detected
• Week-over-week spend delta tracking
• Auto-suggest memory reduction via actor settings API
• Export to Google Sheets