Crawl: the most reliable web crawler APISee what’s new
Zenrows
Talk to sales Start free

What shipped, and when.

Every change a Zenrows user would notice, newest first. What it does, and where to find it in the app.

9 Oct 2026

Crawl: the most reliable web crawler API

Crawl

Give Crawl any public URL and get back every page behind it. It follows the links as deep as you choose, drops duplicates, fetches every page past anti-bot protection, and returns every URL it found, plus each page's HTML if you ask.

Read the postOpen in the app

5 Oct 2026

AI agents can buy Zenrows themselves

Billing

AI agents can now buy and call Fetch and Extract at agents.zenrows.com, with no Zenrows account or API key. They read the live prices, get a free agent credential, and pay over x402, MPP or a card. Credits draw at the same weights as your account, and only successful requests are charged.

Read the docs

2 Oct 2026

Set a credit cap on each API key

Billing

Give a key its own limit per billing period, day, week or month. When it reaches the cap only that key stops, your other keys keep working, and you are alerted at 80% and 100%.

15 Sep 2026

Use Zenrows inside Dify

Integrations

Install the Zenrows plugin from the Dify Marketplace and your Dify workflows and agents get Fetch, Extract and Batch as ready-made tools. Paste your API key once, and there is no HTTP node to configure.

Read the docs

Ask for a site, and Extract prepares it

Extract

You no longer wait for us to enable a domain. Submit it with a few example URLs and Extract prepares it, usually in minutes. It runs on a monthly allowance and costs no credits.

Open in the app

8 Sep 2026

See which sites Extract covers

Extract

One request returns every prepared domain, so you can check coverage before you build, and ask for anything missing.

Open in the app

7 Sep 2026

Rerunning failed tasks returns straight away

Batch

Retrying only the failed tasks in a run responds immediately instead of blocking on the rest of the job.

Open in the app

24 Aug 2026

Run Extract inside a Batch job

Batch

Set extract on the job or per task and results come back as structured JSON, not just pages. Batch still adds no charge of its own: you pay Extract's own price.

Open in the app

21 Aug 2026

Choose which fields Extract sends back

Extract

Pick the fields you actually use instead of taking the default set, so the response carries what you asked for and nothing else.

Open in the app

18 Aug 2026

Export a whole run in one file, up to 100 GiB

Batch

Bundle every result from a run into one archive instead of collecting them one at a time. Each download link is valid for 2 hours; reading the export again within 12 hours issues a fresh one. Results stay for 14 days.

Open in the app

31 Jul 2026

The Zenrows CLI

CLI

Fetch a page, extract structured data, run a Batch job or drive a browser from the terminal. No API key to start: the first call provisions a Free account, and running zenrows account claim makes it yours.

Read the docs

Activity Log: every request you've made

Analytics

One feed of your Fetch, Extract, Batch and Browser activity. Filter by product, search by URL, pick a window from 24 hours to 60 days, and open a row for the full request.

Open in the app

29 Jul 2026

POST targets, and send your own headers

Batch

A task can carry an HTTP method and a request body, stored verbatim and sent as-is, up to 16 KiB, plus custom headers set per job or per task.

Open in the app

22 Jul 2026

Put a job on a schedule

Batch

Run a job once, on an interval, or on chosen days at chosen times in your timezone.

Open in the app

18 Jul 2026

Add URLs to a job that is already running

Batch

A job doesn't have to be complete before it starts. Keep adding targets while it works, 10,000 per call, and close it when you have nothing left to send.

Open in the app

16 Jul 2026

Extract is live: structured JSON from one parameter

Extract

Add extract=auto to a request and the response carries the raw HTML and the page's data as clean JSON. No selectors to write, no schema to define, nothing to fix when the site changes. Currently free.

Open in the app

Batch is live: send a list, collect the results

Batch

Submit up to 100,000 URLs as one job and pick the results up when it finishes, instead of holding a connection open per URL. Up to three jobs run at once (Crawl jobs count too), and results stay for 14 days.

Read the postOpen in the app

Built in the open.

The CLI and the MCP server are open source. Start with either.

Start building free 5,000 credits/month, always.