Core concepts
These are the words you will meet in the Console, the API and these docs. Each entry links to the page that covers it in depth.
Platform
| Term | Meaning |
|---|---|
| Actor | A program packaged as a Docker image that scrapes or automates something. It has an input schema (which generates its input form), a README, versions and optional pricing. Actors are named username~name, for example lantfly~web-scraper. See Actors. |
| Store | The catalog of public actors you can run, with ratings and a quality score. |
| Version | A numbered release of an actor (such as 1.2) that points at an image tag. It carries its own input, dataset and output schemas, environment variables and standby settings. One version is the default; new runs use it unless they name another build. |
| Build | Building an actor image from a Git repository or from files pushed with the CLI. A successful build adds a version and, by default, makes it the default. |
| Run | One execution of an actor with an input. A run has a status (READY, RUNNING, SUCCEEDED, FAILED, TIMED-OUT, ABORTED), a log, its own storages and its usage. See Runs. |
| Run token | A short-lived token Lantfly gives each run. The actor uses it to call the API for its own run, its own storages and the named storages of the account. It stops working shortly after the run ends. |
| Task | An actor saved with an input and run options, so you can run it again in one click, from a schedule or from the API. See Tasks. |
| Dataset | A list of result items (JSON objects) that a run pushes. Export it as JSON, CSV, Excel and more. See Datasets. |
| Key-value store | Records stored by key: the run's INPUT, its OUTPUT, files, screenshots, state. See Key-value stores. |
| Request queue | A queue of URLs (requests) to crawl, de-duplicated by a unique key. See Request queues. |
| Named storage | A dataset, key-value store or request queue with a name. Named storages are kept until you delete them; a run's unnamed storages are removed after your plan's retention period. |
| Schedule | A cron expression and time zone that starts actors or tasks automatically. See Schedules. |
| Monitor | A rule checked when a run finishes (failed, too slow, too few results, empty fields, too costly) that sends an alert. See Monitors. |
| Webhook | An HTTP POST Lantfly sends to your URL when a run event happens, signed with a secret. See Webhooks. |
| Proxy | Lantfly's rotating proxy: one endpoint in front of several proxy groups, with country selection and sticky sessions, billed per GB. See Proxy. |
| Standby | A mode where an actor keeps a warm run that answers HTTP requests sent to https://lantfly.com/api/v3/standby/:actor/..., started on demand. It is turned on per actor version. |
| Live view | A web page an actor serves while it runs, shown in the run's Live view tab. |
| Organization | A team account: several people share actors, runs, storages and billing, each with a role. Requests act for an organization when they send the X-Lantfly-Account header or use a token created for it. See Organizations. |
| API token | A secret that lets your code call the API for your account. It can be read-only, limited to some actors or limited to some permissions. See API tokens. |
Usage and billing
| Term | Meaning |
|---|---|
| Usage | What your runs consume, in USD: compute units, paid results and events, and proxy traffic. |
| Compute unit (CU) | 1 GB of memory for 1 hour. A run with 2 GB for 30 minutes uses 1 CU. |
| Pay per result / pay per event | Pricing an actor developer can set: a price per dataset item (result_item) or per named event the actor charges. Prices can differ by plan. You can cap them per run with Max items and Max cost per run. |
| Plan | Your monthly (or yearly) subscription. Each plan includes some usage per month and sets limits such as concurrent runs and how long run storage is kept. |
| Credit | Prepaid USD in your account's wallet, used after your plan's included usage. It does not expire. |
| Overage | Usage above included usage and credit on a paid plan, billed after the period, up to your spending limit. |
See Billing for prices and plan limits.
Listening
| Term | Meaning |
|---|---|
| Listening project | A topic you follow, such as a brand, a campaign or a competitor. It holds sources, collected mentions, analytics, alerts and reports. |
| Source | One thing a project watches: a hashtag, a profile, a page, a keyword, a product, a shop, a place or an app on a given platform. Each collection of a source is a run. |
| Mention | One post, comment or review a project collected. Mentions carry the text, the author, engagement numbers, sentiment and media. |
| Creator | An author whose posts appear in your projects, with followers, engagement and engagement rate. The Creator directory gathers creators from all your projects; you can track a creator to follow their profile. |
| Campaign | A date range, a budget and the hashtags, keywords or creators of a marketing campaign inside a project. Lantfly works out its posts, reach, engagement, CPM, CPE and ROI. |
See Listening for the full guide.