Aaaaand we just launched Puppet Master, a new serverless API for controlling headless chrome in the cloud.
Our goal is to make it really simple for developers to scale automated browser workflows via Puppeteer.
- Simple: Quick to setup and integrate
- Standard: Just HTTPS + JSON
- Managed: Hosted via serverless functions
- Scalable: Scales to any capacity
- Usage based: Pay only for what you use
- Focused: We only do one thing and do it well
- Immutable deployments: All API updates are optional
Puppeteer is an open source tool released by Google to automate headless Chrome. Puppet Master allows you to run common Puppeteer workflows in the cloud.
Navigates to a page and captures a screenshot via Puppeteer's Page.screenshot.
All Puppeteer options are supported.
Returns an image as either
image/jpegdepending on the
Navigates to a page and captures a PDF via Puppeteer's Page.pdf.
All Puppeteer options are supported.
Returns a PDF as
In addition to the standard parameters, all functions optionally accept a set of additional common options.
- viewport - Set the browser window's viewport dimensions and/or resolution.
- Useful to trick websites into displaying different content or styles based on an emulated device.
- gotoOptions - Customize the
- Useful when you need to customize behavior around when Puppeteer considers a page navigation "finished".
- emulateDevice - Make it look like the screenshot was taken on the specified device.
- Use the
nameproperty from one of the built-in devices.
Try setting viewport.deviceScaleFactor to
2to emulate a retina display.
The differences can be broken down into two areas:
- We use serverless functions to provide maximum scalability while minimizing cost.
- This also means that you only pay for what you actually use.
- You could set this up yourself, but headless chrome is a resource hog and who wants to setup and manage another async task queue?
- We also allow you to run functions locally which is essential for debugging.
The core motivation behind this project lies in trying to tackle the problem of open source sustainability.
Towards that end, every aspect of this SaaS product aside from the core functionality was generated automatically via Saasify, the Shopify for SaaS. That includes the template-based marketing website, OpenAPI-based API docs, serverless function hosting, user account management, usage-based billing integration, legal docs, and support.
It's our hope that by greatly simplifying and automating the process of launching SaaS APIs, open source authors will be much more inclined to monetize their work and help to sustain their passion while still holding true to the underlying nature of OSS.
Puppet Master is therefore a SaaS product that is fully generated from the source code in this repository. If you know how to create an open source TypeScript project like this one, you're already 95% of your way towards creating your own monetizable SaaS product!
We're actively looking for feedback and feature suggestions.
All APIs come with a free public tier that is rate-limited so you can test things out before signing up for a subscription.
Saasify uses Stripe to securely handle all payments, subscriptions, and invoices.
We use usage-based, metered subscriptions that charge fractions of a cent based on the number of authenticated API calls
Signing up for a subscription starts a monthly billing cycle. Note that you will only be charged for what actually you use.
See pricing for more details.
Collaboration is what makes open source truly powerful! If you're interested in working on Puppet Master or have an idea for your own Saasify product, please contact us to learn more. Thanks!
If your API calls are not working or returning unexpected results, the best way to troubleshoot is to try and reproduce your issue by running the project locally.
- Clone this repository
npm install -g saasify
saasify devat the repo root specifying an additional environment variable
PUPPET_MASTER_DEBUG=1 saasify dev
This will run the deployment locally with a non-headless browser and
Then just change the URL you're calling to be the corresponding
localhostURL and investigate the issue.