You can see what an AskEdgar API query will cost before you spend anything. The free estimate endpoint needs no account or API key, works in a browser, and returns a projected cost for any data query.
This guide shows how pricing works, how to run and read an estimate, how to size bigger jobs like a daily watchlist, and how to check what a call actually cost afterward.
How AskEdgar API pricing works
The API uses prepaid credits, with no subscription or monthly commitment. According to the Pricing & Credits page:
- You pay for data, not requests. Each call is charged by the volume of data it returns, priced per GB. A small response costs less than a large one from the same endpoint.
- Rates vary by endpoint. Harder-to-source data, such as dilution data, offerings and fund ownership, costs more than simpler lookups.
- Add credits before you request data. You buy them on the API platform's billing page.
Estimate a query in your browser
The estimate endpoint is a normal web address, so you can try it without writing any code.
- Pick the endpoint you plan to call. For example, dilution-rating or dilution-data. The API reference lists them all.
- Pick your tickers. Use real symbols, separated by commas, up to 100 per estimate.
- Build the address. Put both into the estimate URL: https://eapi.askedgar.io/estimate?endpoint=dilution-rating&ticker=DVLT
- Open it in your browser. You'll get back a short JSON response with the projected cost. No key or sign-in needed.
To estimate the same query for several tickers, list them all: https://eapi.askedgar.io/estimate?endpoint=dilution-data&ticker=DVLT,XXII
Reading the estimate
A single-ticker dilution-rating estimate returns these fields:
- records_returned: how many records the query is expected to return.
- estimated_bytes: the projected size of the response. This drives the cost.
- estimated_cost_dollars: the projected cost in US dollars. estimated_cost_microdollars is the same figure in millionths of a dollar, which is easier to add up in code.
- price_per_gb_dollars: the rate used for that endpoint.
- supports_date_range: whether the endpoint accepts date_from and date_to. If it's false, date filters won't change the estimate.
- limit, date_from, date_to: echo back the filters you passed, so you can confirm the estimate matches the call you plan to make.
Treat it as a projection, not a quote. The docs say estimates are statistical and the actual cost can land meaningfully above or below them. Prices can also change, so run a fresh estimate rather than reusing an old number.
Estimating bigger jobs
The estimate endpoint takes a few more options that let you size real workloads. These come from the Estimate Usage reference.
- A specific watchlist (most accurate). List your exact tickers with the ticker parameter, up to 100 per estimate. The docs call this a tight estimate, because it measures real data for those tickers.
- A broad screen (rough guide). Use the ticker_count parameter with a number from 1 to 10,000 instead of naming tickers. It averages over a random sample, so treat the result as directional. Use ticker or ticker_count, not both.
- Match your limit. If your real call uses a limit, pass the same limit to the estimate, from 1 to 100. Otherwise the estimate assumes no cap.
- Narrow the dates. For endpoints that support it, add date_from and date_to in YYYY-MM-DD format. Narrower ranges make the estimate both smaller and more reliable.
For a recurring job, multiply. An estimate covers one call. A watchlist you refresh every trading day costs roughly that estimate times the number of runs, so a quick multiplication tells you what to budget per month. (This is our rule of thumb, not a figure from the docs.)
Cut cost on dilution data. On dilution endpoints, adding show_in_play=true to your actual request returns only instruments that pass a screen on remaining size and price relative to the market price, which the docs say can significantly reduce cost. It doesn't check maturity or expiration dates, so it can leave out instruments you'd still want to review.
After you query: what it actually cost
You don't have to guess afterward either. Every paid response tells you what it cost.
- The usage block. Each response ends with a usage section. It includes cost_dollars for that call, credits_remaining_dollars for your balance, and response_bytes and billed_bytes for the size. Comparing cost_dollars with your estimate shows how close the projection was for your workload. Data endpoints return 10 records per page, while an estimate covers every record for the ticker, so add up cost_dollars across all pages before comparing.
- The X-Credits-Remaining header. Billed responses also carry your remaining balance in US dollars as a response header, which is handy for logging or alerts.
- Running low. If your balance can't cover a request, the API returns HTTP 402 with the code insufficient_credits, your current balance and the request's estimated cost. The docs say the request is blocked before any data comes back and that you're never charged for a failed request.
A simple habit for any tool you build: estimate before adding a new query to a schedule, then watch cost_dollars for the first few runs.
Common questions
Does running an estimate cost anything? No. The estimate endpoint is free and needs no API key.
Do I need an account to estimate? No. You only need an account, a key and credits to request actual data.
How accurate is an estimate? It's a projection based on recent data. Naming specific tickers gives a tight estimate, while ticker_count gives a rougher one. Actual cost can come in above or below.
Am I charged for errors? The docs say you're never charged for a failed request, and requests blocked for low balance return a 402 before any data is sent.
Can my AI coding assistant check costs? Yes. AskEdgar's MCP server lets an assistant call the free estimate endpoint before it writes code that requests data.
Can I use the data in a product for my customers? Standard API access is licensed for personal use. For business, institutional or customer-facing use, or redistribution, email support@askedgar.io about a license.
Start building
Estimate the queries your tool needs, then get your API key, add credits and follow the Quick Start to make your first request. Choosing a data provider? Start with "How to Choose a Small-Cap Stock API for Dilution Research".
Last checked against the AskEdgar docs on October 21, 2026.