OCaml Packages Scraper

Returns: one row per OCaml package version with name, version, synopsis, license, authors, dependencies, checksums, publishedAt. Input: {"searchQueries": ["yaml"], "maxItems": 50} Best for: dependency audits of OCaml projects, license and supply-chain checks, tracking new releases of opam packages.

{
    "name": "lwt",
    "version": "6.1.2",
    "isLatest": true,
    "synopsis": "Promises and event-driven I/O",
    "license": "MIT",
    "publishedAt": "2026-04-29",
    "installCommand": "opam install lwt",
    "usedByCount": 511
}

What does OCaml Packages Scraper do?

OCaml Packages Scraper turns the package pages of ocaml.org into clean JSON. ocaml.org has no public per-package JSON API, so this Actor works as an OCaml package API: give it search queries (the same ones you would type into the ocaml.org package search, including filters such as author:"Jérôme Vouillon") or a list of opam package names and URLs, and it returns one structured record per package version.

Each record holds the package name and version, whether that version is the latest, the one-line synopsis and the longer description, the license, homepage, opam file link and number of published versions, the authors and maintainers, the source archive URL with its checksums, the date the version was added to opam-repository, the install command, and the full dependency lists (runtime and development) with their version constraints. It also counts how many packages depend on this one.

The Actor reads public pages only. No login, no account and no API key for ocaml.org are needed. It is useful for dependency audits, license reviews, software bills of materials, release monitoring, and building datasets of the OCaml ecosystem.

What data can you get from ocaml.org?

Field	Description	Example
id	Stable record ID: name@version	lwt@6.1.2
name	opam package name	lwt
version	Package version	6.1.2
isLatest	True when this is the newest version	true
synopsis	One-line summary of the package	Promises and event-driven I/O
description	Longer description, paragraphs separated by a blank line; null when there is none	A promise is a value that may become determined in the future.
license	License from the license link, without the word "License", or the first sentence of the README License paragraph; null when both are absent	MIT
homepage	Project homepage or repository	https://github.com/ocsigen/lwt
opamFileUrl	The opam file in opam-repository	https://github.com/ocaml/opam-repository/blob/master/packages/lwt/lwt.6.1.2/opam
versionsCount	Number of published versions	32
authors	Author names	["Jérôme Vouillon", "Jérémie Dimino"]
maintainers	Maintainer names, sometimes with an email	["Raphaël Proust <code@bnwr.net>", "Anton Bachin"]
sourceArchiveUrl	Download URL of the release source archive	https://github.com/ocsigen/lwt/archive/refs/tags/6.1.2.tar.gz
checksums	Archive checksums as algorithm=hex	["md5=00bc7eb2772a28530f7365a7040f0c4f", "sha512=a473..."]
publishedAt	Date added to opam-repository (YYYY-MM-DD)	2026-04-29
installCommand	opam command that installs this version	opam install lwt
dependencies	Dependencies as {name, constraint} objects	[{"name": "dune", "constraint": ">= \"3.18\""}]
devDependencies	Test, doc and dev dependencies as {name, constraint}; empty when none	[{"name": "odoc", "constraint": "with-doc & >= \"2.3\""}]
dependenciesCount	Number of dependencies	5
usedByCount	Number of packages that depend on this one; 0 when none	511
docsUrl	Package documentation on ocaml.org	https://ocaml.org/p/lwt/latest/doc/index.html
packageUrl	The package page that was scraped	https://ocaml.org/p/lwt/latest
searchQuery	The search query that found the package; null for packages input	http
scrapedAt	Time of the scrape (ISO 8601, UTC)	2026-10-15T19:08:56.393Z
sourceUrl	Page the record came from	https://ocaml.org/p/lwt/latest

How to use OCaml Packages Scraper

1. Open OCaml Packages Scraper in Apify Console and go to the Input tab.
2. Type one or more search queries (for example yaml, http or author:"Jérôme Vouillon"), or list packages by name or URL (for example lwt, yojson@2.1.0 or https://ocaml.org/p/yojson/2.1.0). You can use both fields in the same run.
3. Set Maximum results to the number of packages you want in total.
4. Click Start. Each search query is paged through the ocaml.org package search, 50 results per page, and every result's package page is opened and parsed.
5. When the run finishes, open the Output tab, or export the dataset as JSON, CSV, Excel, XML or RSS.

Tips:

A bare package name always gives the latest version. To get an older release, use name@version, name.version or a full version URL such as https://ocaml.org/p/yojson/2.1.0; that record then has isLatest set to false.

maxItems is one budget shared by every target in the run. Packages listed under packages are scraped first, then the search queries in order, so a large first search can use the whole budget. If the same package version turns up twice (in two searches, or as a package and a search result), it is returned once.

The Actor keeps a modest pace (about two requests per second) so a run of 50 packages usually finishes in well under a minute.

Input

Field	Type	Required	Description
searchQueries	array of strings	One of the two	ocaml.org package search terms, free text or filters such as author:"Name". Every result's package page is scraped. Use * to walk the whole package index (set maxItems to bound it).
packages	array of strings	One of the two	opam package names (lwt), name@version (yojson@2.1.0) or ocaml.org package URLs (https://ocaml.org/p/yojson/2.1.0).
maxItems	integer	No	Maximum number of packages across all targets, 1-10000, default 50.
proxyConfiguration	object	No	Proxy settings. Default: no proxy (ocaml.org works without one).

Give at least one search query or one package. A search URL copied from ocaml.org (https://ocaml.org/packages/search?q=yaml) also works in either field.

{
    "searchQueries": ["yaml", "author:\"Jérôme Vouillon\""],
    "packages": ["lwt", "https://ocaml.org/p/yojson/2.1.0"],
    "maxItems": 50,
    "proxyConfiguration": { "useApifyProxy": false }
}

Output

Each package version is one item in the run's dataset. These two items come from a real run with {"packages": ["lwt", "https://ocaml.org/p/yojson/2.1.0"]} (the lwt description is shortened here):

[
    {
        "id": "lwt@6.1.2",
        "name": "lwt",
        "version": "6.1.2",
        "isLatest": true,
        "synopsis": "Promises and event-driven I/O",
        "description": "A promise is a value that may become determined in the future.\n\nLwt provides typed, composable promises. ...",
        "license": "MIT",
        "homepage": "https://github.com/ocsigen/lwt",
        "opamFileUrl": "https://github.com/ocaml/opam-repository/blob/master/packages/lwt/lwt.6.1.2/opam",
        "versionsCount": 32,
        "authors": ["Jérôme Vouillon", "Jérémie Dimino"],
        "maintainers": ["Raphaël Proust <code@bnwr.net>", "Anton Bachin"],
        "sourceArchiveUrl": "https://github.com/ocsigen/lwt/archive/refs/tags/6.1.2.tar.gz",
        "checksums": [
            "md5=00bc7eb2772a28530f7365a7040f0c4f",
            "sha512=a4737a2fc44aaeedca46ec6134c07013d5cf44039a988a6e79a40715a53e43601d4c91bbf11a8e89540528d912b8fa487e3068465c136c1156a872258eb3443b"
        ],
        "publishedAt": "2026-04-29",
        "installCommand": "opam install lwt",
        "dependencies": [
            { "name": "ocplib-endian", "constraint": null },
            { "name": "dune-configurator", "constraint": null },
            { "name": "cppo", "constraint": "build & >= \"1.1\"" },
            { "name": "ocaml", "constraint": ">= \"4.14\"" },
            { "name": "dune", "constraint": ">= \"3.18\"" }
        ],
        "devDependencies": [
            { "name": "odoc", "constraint": "with-doc & >= \"2.3\"" },
            { "name": "ocamlfind", "constraint": "dev & >= \"1.7.3-1\"" }
        ],
        "dependenciesCount": 5,
        "usedByCount": 511,
        "docsUrl": "https://ocaml.org/p/lwt/latest/doc/index.html",
        "packageUrl": "https://ocaml.org/p/lwt/latest",
        "searchQuery": null,
        "scrapedAt": "2026-10-15T19:08:56.393Z",
        "sourceUrl": "https://ocaml.org/p/lwt/latest"
    },
    {
        "id": "yojson@2.1.0",
        "name": "yojson",
        "version": "2.1.0",
        "isLatest": false,
        "synopsis": "Yojson is an optimized parsing and printing library for the JSON format",
        "description": "Yojson is an optimized parsing and printing library for the JSON format.\n\nydump is a pretty-printing command-line program provided with the\nyojson package.",
        "license": "BSD-3-Clause",
        "homepage": "https://github.com/ocaml-community/yojson",
        "opamFileUrl": "https://github.com/ocaml/opam-repository/blob/master/packages/yojson/yojson.2.1.0/opam",
        "versionsCount": 15,
        "authors": ["Martin Jambon"],
        "maintainers": ["paul-elliot@tarides.com", "nathan@tarides.com", "marek@tarides.com"],
        "sourceArchiveUrl": "https://github.com/ocaml-community/yojson/releases/download/2.1.0/yojson-2.1.0.tbz",
        "checksums": [
            "sha256=9fcb1ff2db58ab259f9228796b0ada4794eae97177b1833371380c4e4f90b15d",
            "sha512=31ab8580e0e4e8a05459fbec2fb9424a3df4cf5aef6df21fc577dbe3d32e81103a632c6a511c4dcf1f51819e88b38d2ce7d6888545bc3f359e69e070f8d2e69c"
        ],
        "publishedAt": "2023-05-02",
        "installCommand": "opam install yojson.2.1.0",
        "dependencies": [
            { "name": "seq", "constraint": ">= \"0.2.2\"" },
            { "name": "cppo", "constraint": "build" },
            { "name": "ocaml", "constraint": ">= \"4.02.3\"" },
            { "name": "dune", "constraint": ">= \"2.7\"" }
        ],
        "devDependencies": [
            { "name": "odoc", "constraint": "with-doc" },
            { "name": "alcotest", "constraint": "with-test & >= \"0.8.5\"" }
        ],
        "dependenciesCount": 4,
        "usedByCount": 287,
        "docsUrl": "https://ocaml.org/p/yojson/2.1.0/doc/index.html",
        "packageUrl": "https://ocaml.org/p/yojson/2.1.0",
        "searchQuery": null,
        "scrapedAt": "2026-10-15T19:08:56.963Z",
        "sourceUrl": "https://ocaml.org/p/yojson/2.1.0"
    }
]

Fields that a package page doesn't show are null (single values) or an empty list (lists), so every item has the same shape. For example, some packages have neither a license link nor a README License paragraph, or have no description.

How much does it cost to scrape ocaml.org?

OCaml Packages Scraper is charged per result. You pay only for the package records returned, and maxItems caps how many that can be. Searches that match nothing, unknown packages and failed pages cost nothing. See the Pricing tab for the current price.

Use with AI agents (MCP)

Connect OCaml Packages Scraper to Claude, Cursor or any MCP client through the Apify MCP server:

{
    "mcpServers": {
        "apify": {
            "url": "https://mcp.apify.com/?tools=axiomworks/ocaml-packages-scraper"
        }
    }
}

Example prompts:

1. "Use the OCaml packages scraper to get lwt, cmdliner and yojson, and list each one's license, latest version and dependencies with constraints."
2. "Search ocaml.org for packages by author:"Jérôme Vouillon" and make a table of name, version, publish date and how many packages use each one."
3. "Find the 20 most relevant OCaml HTTP packages and tell me which have no license value or have not had a release since 2024."

An agent sees the input fields searchQueries, packages and maxItems, so it can search or look up exact packages in one call.

Integrations and API

Run OCaml Packages Scraper from your own code with the Apify API. Set APIFY_TOKEN to your Apify API token (send it in the Authorization header, never in the URL).

Python

import os
from apify_client import ApifyClient

client = ApifyClient(os.environ["APIFY_TOKEN"])
run = client.actor("axiomworks/ocaml-packages-scraper").call(
    run_input={"packages": ["lwt", "yojson@2.1.0"], "searchQueries": ["yaml"], "maxItems": 50}
)
for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(item["id"], item["license"], len(item["dependencies"]))

JavaScript

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: process.env.APIFY_TOKEN });
const run = await client.actor('axiomworks/ocaml-packages-scraper').call({
    searchQueries: ['yaml'],
    maxItems: 50,
});
const { items } = await client.dataset(run.defaultDatasetId).listItems();
for (const p of items) {
    // p.name, p.version, p.isLatest, p.dependencies ...
}

cURL

curl -X POST "https://api.apify.com/v2/acts/axiomworks~ocaml-packages-scraper/run-sync-get-dataset-items" \
  -H "Authorization: Bearer $APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"searchQueries": ["yaml"], "maxItems": 20}'

The Integrations tab connects the Actor to Make, Zapier, n8n, Google Sheets, Slack and webhooks, so you can, for example, send a row to a spreadsheet for every package version in a weekly scheduled run.

FAQ

Is there an official ocaml.org JSON API for packages? Not for package metadata. The package pages are HTML, so this Actor reads them and returns the same data as structured JSON. Think of it as an OCaml package API you can call from code or from an AI agent.

How do I get an older version of a package? Use name@version (for example yojson@2.1.0) or the version's URL (https://ocaml.org/p/yojson/2.1.0) in packages. A bare name and every search result give the latest version.

Which search queries work? Anything the ocaml.org package search accepts: free text such as json or http server, and filters such as author:"Anil Madhavapeddy". Non-ASCII names work as typed.

Why did my run return 0 results? The search matched no packages, or the package name doesn't exist on ocaml.org (HTTP 404). The run still ends as succeeded, and the log and status message say which target found nothing. Check the spelling, or try the search on ocaml.org first.

Why is license or description null for some packages? License uses the sidebar license link first, then the first sentence or short text of the paragraph immediately after the README License heading. Some package pages have neither source, or no description section. The Actor returns null instead of guessing. The opam file (opamFileUrl) is the place to check those packages by hand.

How fresh is the data? Every run reads the live pages, so the data is as current as ocaml.org at the time of the run (scrapedAt). Schedule the Actor to track new releases.

Do I need a proxy? Usually not. ocaml.org answers normal requests without one, so the default is no proxy. If you see HTTP 403 or 429 in the log, turn on Apify Proxy; blocked requests are retried on a new session automatically.

How many packages can one run return? Up to 10,000 (maxItems). Large searches are paged 50 results at a time until the search runs out or maxItems is reached.

Is it legal to scrape ocaml.org?

This Actor collects only publicly available package metadata that anyone can see on ocaml.org without logging in. Package pages list author and maintainer names and sometimes email addresses; if you store or process those, you are responsible for doing so lawfully (for example under GDPR) and for respecting the site's terms. This is not legal advice.

OCaml Packages Scraper is an independent tool. It is not affiliated with, endorsed by or sponsored by ocaml.org or the OCaml Software Foundation.

Feedback

Found a bug or need a field that isn't here? Open an issue on the Issues tab and we'll look at it.