Usage, limits and retention
Understand the AIHackr 20-page free trial, crawl reservations, request limits, seven-day result retention and current service availability.
On this page
Current free trial
Each verified email account receives 20 lifetime trial pages. You do not need a card. Sessions, API keys and MCP connections share that allowance; creating another key or signing in again does not reset it. Paid subscriptions and payment collection are not live.
| Limit | Current behavior |
|---|---|
| Page operations | Scrape, Map and Extract each reserve one page. |
| Crawl accounting | Reserves the entire requested page maximum, even when fewer pages finish or an upstream request fails. |
| Customer daily allowance | At most 20 reserved pages per day, within the lifetime trial. |
| Service safety allowance | 1,000 attempted pages per day shared across the deployment. |
| Concurrency | Three requests and two active crawls per workspace. |
| Engine request rate | 30 operations per minute per workspace. |
| Crawl bounds | 100 pages and depth 5, subject to account allowance. |
| Result retention | Seven days for stored crawl jobs and results. |
| Retained crawl page content | Up to 60,000 characters per requested format; inspect truncation warnings. |
Plan a bounded run
Start with a single scrape. For a crawl, explicitly set limit to a small number; the API default of 20 can reserve a new account’s full allowance. Save results externally if you need them after seven days. Avoid overlapping crawls and use deliberate backoff on rate limits.
Current availability
Scrape, page-level Map, asynchronous Crawl, structured Extract, email accounts, customer API keys and the eight browser/crawl MCP tools are available. Web search, batch scraping, live subscriptions and hosted OAuth are not enabled. Consult the live capabilities endpoint for deployment state.
Public targets and access
Only public HTTP(S) sites are accepted. Private addresses, embedded credentials and custom ports are rejected. Robots rules and the target’s response can prevent processing. Logged-in target sessions, CAPTCHA solving and anti-bot bypass are not offered. Rendering does not guarantee access or a complete result.