Keepa API access is metered by tokens. Every API call consumes tokens, and your token balance refills every minute based on your Keepa subscription plan. This page covers how tokens work, what they cost, and how Agellic manages them for you.
Minimum Plan Requirement
Agellic requires a Keepa plan with at least 20 tokens per minute. The standard 1 token/minute plan will not work. A single search takes 11 tokens, meaning you would wait 11 minutes for one set of results. A conversational interface needs fast responses to be useful.
The 20 tokens/min plan is all you need to use Agellic effectively. Higher plans process large batch operations faster, but are not required.
You can check available plans at keepa.com.
Token Costs by Operation
Finder Searches
Finder searches let you discover products matching specific criteria. The base cost is 10 tokens plus 1 token per 100 ASINs in the result set. Search Insights (used by the AI agent to refine searches iteratively) adds 30 tokens.
- Standard search (up to 100 results): 11 tokens
- Larger search (up to 1,000 results): 20 tokens
- Search Insights: +30 tokens on top of the search cost
Product Lookups
Cost depends on the level of detail requested:
- Quick lookup (basic price, BSR, reviews): 1 token per ASIN
- Enriched lookup (full detail with offers, stock, Buy Box, ratings): 9 tokens per ASIN
- Keepa chart image: 1 token per ASIN
Enriched lookups are the default when you ask the agent to look up a product. They fetch everything needed for advanced analytics. Quick lookups are used when you only need the basics.
Saved Product List Refreshes
Refreshing a saved product list is the most token-intensive operation because every ASIN gets an enriched lookup:
- 10-product list refresh: 90 tokens
- 50-product list refresh: 450 tokens
- 100-product list refresh: 900 tokens
On a 20 tok/min plan, a 100-product refresh will take time as tokens refill. On a 60 tok/min plan, the same refresh completes much faster. Agellic runs these as background jobs so you can keep working while they process.
Keep saved lists focused. A 50-product list of strong candidates is more useful, and cheaper to refresh, than a 500-product list you haven't filtered yet.
How Agellic Manages Your Tokens
Token Reserve
Agellic maintains a 200-token reserve for interactive use. Background jobs (saved search refreshes, list refreshes) only consume tokens above this reserve. This ensures you always have tokens available for live agent conversations, even while background jobs are running.
Smart Queuing
When a job requires more tokens than are currently available, Agellic queues it and processes it as tokens refill. You can keep chatting while jobs run in the background. Jobs continue processing even if you close your browser.
Caching
Product data is cached for 24 hours. If you look up the same ASIN twice in one day, the second lookup is free and serves directly from cache. This significantly reduces token consumption for iterative research where you revisit products.
Balance Checking
You can check your token balance at any time by asking the agent: "What's my token balance?" or "How many tokens do I have?"