What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
To turn HTML and CSS into an image through an API, send the markup to a hosted renderer and request a PNG or PDF; to capture an existing page, give the service a publicly reachable URL. A hosted API avoids running browsers yourself, while Playwright and Puppeteer offer finer browser control if your team can operate the browser infrastructure. Your choice depends on the input you have, how precisely you need to control rendering, and who will handle retries, scaling, and browser failures.
Choose an input path before choosing an API
HTML-to-image services generally accept one of three inputs: raw HTML and CSS, a public URL, or structured data for a named template. Those are different workflows, not interchangeable request formats. Use raw markup when your application owns the page content, a URL when the page already exists on the public web, and a template endpoint when the rendering service stores a reusable design that your application fills with data.
For example, html2img documents all three approaches: POST https://app.html2img.com/api/html for HTML and CSS, POST https://app.html2img.com/api/screenshot for a publicly accessible URL, and POST https://app.html2img.com/api/v1/templates/[slug] for JSON sent to a named template. Its GET https://app.html2img.com/api/me endpoint returns account status without consuming a credit. See the html2img official documentation for its current request formats and client libraries.
Raw HTML and CSS
Send a complete HTML document or a fragment, plus styles, to be rendered by a browser-backed service. This is a direct fit for invoices, certificates, cards, and other content your application generates. If your markup runs JavaScript to populate or modify the page, confirm that the chosen endpoint executes inline JavaScript and understand how it signals that rendering is finished.
Free tools Windows power users keep installed
One-click scans. No signup required.
#1 Best Overall
- HTML CSS Design and Build Web Sites
- Comes with secure packaging
- It can be a gift option
Public URL
A URL-based capture asks the renderer to navigate to a page as a browser would. The page must be reachable from the service; localhost, a private network address, or a page behind an unprovided login is not automatically accessible. Pages that load data after the initial HTML response may need a selector wait, delay, or asynchronous completion mechanism.
Structured template data
A template API separates layout from content: your application sends JSON values, while the service renders the named template. This can make repeated output consistent, but ties rendering to the template system and its input schema. Validate data against the template’s expected fields rather than assuming arbitrary HTML is accepted by that endpoint.
Use a hosted API for a managed browser workflow
A typical hosted request has four parts: an API key, an input, rendering options, and a response format. For html2img, the official guide requires an API key in the X-API-Key header; its documentation states, “All API requests require authentication using an API key.” The exact fields vary by endpoint, so confirm the request body for the endpoint you use rather than sending screenshot parameters to an HTML or template route without checking its schema.
The html2img parameter reference documents dimensions, full-page capture, DPI, injected CSS, selector waits, delays, webhooks, and output format. It lists width and height values from 1 to 5000 pixels, and documents PNG or PDF output. For ordinary HTML renders, the guide recommends synchronous requests; for slow URL screenshots, it recommends webhooks. It also recommends DPI 1 for most cases because higher DPI increases processing time and memory use.
Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Scan for outdated or missing drivers - takes under a minuteDriver Scan →The guide lists maintained PHP, Laravel, JavaScript, Python, and Ruby clients. For production code, use the client for your language if it covers the endpoint and options you need; otherwise, a standard HTTP client is sufficient. Keep API keys on the server, not in public browser JavaScript, and handle unsuccessful HTTP responses explicitly. The API reference documents HTTP 400 validation errors and 422 errors for templates.
Control dimensions, page extent, and render timing
Correct content can still produce an unusable image if the viewport, capture boundary, or render timing is wrong. Set these deliberately, and validate the resulting output against representative pages rather than assuming a single set of options will work for every document.
Viewport width and height
Width and height determine the layout viewport for a capture. A narrow viewport may trigger a mobile layout; a wider one may reveal desktop columns or change line wrapping. html2img documents a range of 1–5000 pixels for each dimension. Choose values that match the intended output, and remember that changing width can alter content height through reflow.
Viewport capture versus full page
A viewport screenshot captures the visible browser area. Full-page mode is intended to include the page beyond that area. For long pages, check the output dimensions and content: lazy-loaded images or elements that appear only after scrolling can be absent unless the renderer loads them before the capture. Hosted APIs differ in how they implement full-page rendering, so treat the option name as a request, not a guarantee that every site will behave identically.
Wait for content that appears late
Use wait_for_selector when a specific element indicates that the relevant content is ready. A fixed ms_delay is simpler, but it can waste time on fast pages and still be too short on slow ones. For a slow URL capture, html2img documents webhook_url for asynchronous completion; this avoids holding a synchronous request open while the page renders. Webhook consumers should validate incoming requests according to the service’s documented security mechanism and make processing safe to retry.
DPI and PDF output
Use DPI 1 as the html2img guide’s recommended default for most cases. Higher DPI can increase processing time and memory requirements, so reserve it for output that demonstrably needs more detail. If you request PDF, the parameter reference documents scale_to_fit and page-related controls; verify the rendered pagination, margins, and scaling with your actual content before treating the output as print-ready.
When Playwright or Puppeteer is the better fit
Self-hosted browser automation makes sense when the team needs control below the hosted API’s abstraction: a particular browser setup, local files, custom navigation logic, deterministic element masking, or direct access to browser lifecycle and failures. The trade-off is operational responsibility. Your service must provision browser processes and dependencies, manage memory and concurrency, limit runaway jobs, and scale workers as request volume changes.
Playwright
Playwright’s Page API documents page.screenshot() options for PNG, JPEG, and WebP; full-page capture; masking; transparent backgrounds; quality; CSS-pixel or device-pixel scaling; injected styles; and timeouts. These controls are useful when a capture must hide volatile or private elements, match a device scale, or run inside an existing browser automation workflow. Consult the Playwright Page API for the current method signature and option details.
Puppeteer
Puppeteer’s official screenshot guide demonstrates launching a browser, navigating to a page, and calling page.screenshot(). It also shows element capture through ElementHandle.screenshot(). Chrome for Developers describes Puppeteer as a JavaScript library for automating Chrome and Firefox, including screenshots, PDFs, navigation, and UI testing. See the Puppeteer screenshot guide and Chrome for Developers’ Puppeteer overview.
Neither library removes the need to decide when a page is ready. Navigation completion does not necessarily mean client-side data, fonts, or images have finished loading. Wait for the content your capture depends on, set sensible timeouts, and clean up browser contexts and processes even when navigation or screenshot generation fails.
Compare the approaches on the work you need done
| Approach | Input | Capture control | Output | Operational responsibility |
|---|---|---|---|---|
| ScreenshotNeo | URL, with a one-GET screenshot API; also HTML/CSS-to-image capability | 63 options, including viewport, full page, selector, waits, CSS, JavaScript, and request controls | PNG, JPEG, WebP, or PDF | Hosted API; billed only for clean shots |
| html2img | Raw HTML/CSS, public URL, or JSON for a named template | Dimensions, full page, DPI, injected CSS, selector wait, delay, and webhook | PNG or PDF, as documented | Hosted API and account credits; consult vendor terms for current pricing |
| Playwright | Pages controlled by your browser automation code | Browser-level screenshot options including full page, masks, styles, scale, and timeouts | PNG, JPEG, or WebP, as documented | You operate browser dependencies, processes, and scaling |
| Puppeteer | Pages controlled by your browser automation code | Page and element screenshots through its browser automation API | Screenshot formats supported by its current API; see the Puppeteer guide | You operate browser dependencies, processes, and scaling |
Choose by deployment model and input first, then compare the controls you actually need. A hosted service reduces browser operations work but means you depend on its API, options, and usage terms. Running Playwright or Puppeteer gives more direct control but makes browser reliability and capacity your responsibility. The reviewed product documentation does not provide a comparable latency or accuracy benchmark across these options, so it cannot establish which renders fastest or most faithfully on a particular site.
Rank #4
- Brand: Wiley
- Set of 2 Volumes
- A handy two-book set that uniquely combines related technologies Highly visual format and accessible language makes these books highly effective learning tools Perfect for beginning web designers and front-end developers
Or skip the browser setup
For a URL capture, ScreenshotNeo provides a one-GET screenshot API. This cURL example saves the response as WebP; see the ScreenshotNeo API documentation for request options and response details.
The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →curl -G "https://api.screenshotneo.com/v1/shot" -d access_key=YOUR_API_KEY --data-urlencode url=https://stripe.com -o shot.webp
ScreenshotNeo removes cookie and consent banners, newsletter popups, and chat widgets before capture; each cleanup step can be turned off. Bot checks, blank pages, timeouts, failed loads, and cache hits cost nothing, and response headers report the page verdict and billing status. Its MCP server provides screenshot tools for AI agents, and every plan includes all features. The Free plan includes 1,000 screenshots a month with no card; paid plans start at $5 for 3,000 shots. Sign up for ScreenshotNeo’s free plan.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Troubleshoot missing, blank, or incorrect captures
- Authentication or validation error: Check that the API key is sent in the required
X-API-Keyheader for html2img and that the payload matches the endpoint. Its reference documents HTTP 400 for validation errors and 422 for template errors; inspect the response body and correct the field or template data it identifies. - URL cannot be captured: Confirm the page is publicly reachable from outside your network and does not require an interactive login or private VPN. If the page is access-controlled, use an API that supports the necessary documented authentication controls or render the content you control through an HTML input or self-hosted browser.
- Content is missing: A page may render data after initial navigation. Wait for a visible, content-specific selector rather than relying only on navigation, or use a delay when no useful selector exists. Increase a delay only after confirming that timing is the cause.
- Images or lower sections are absent: Check whether the page uses lazy loading and whether full-page mode is enabled and supported. Test the exact page because lazy images can require scrolling or other page interaction before they load.
- Capture is slow or times out: Reduce unnecessary waits, use a selector tied to readiness instead of a long fixed delay, and consider a webhook for slow URL screenshots where the hosted API supports it. In a self-hosted worker, cap concurrent browsers and close contexts after errors.
- Text or layout differs from the expected design: Match the viewport to the intended breakpoint, confirm fonts and other assets are reachable, and check injected styles, device scale, and DPI. Browser rendering can differ if the page receives different resources or viewport dimensions.
- PDF pagination or sizing looks wrong: Review page size, margins, page ranges, landscape orientation, and scaling controls offered by the API. Verify with long and short documents; a setting that fits one page can shrink a longer document unexpectedly.
Plan for reliability, performance, and cost
Hosted services shift browser provisioning and scaling out of your application, but they introduce a network request and vendor-specific usage model. Check the current plan terms and billing unit before estimating recurring costs; the html2img documentation cited here does not establish current price tiers. For either approach, measure your own workload: page complexity, image size, full-page length, waits, and DPI can affect work per capture.
For dependable systems, distinguish request acceptance from completed rendering. A synchronous call is straightforward for a quick render, while an asynchronous webhook fits longer URL jobs. Set client-side timeouts, record the endpoint and options used, and make retry behavior explicit. Retrying a failed capture can be appropriate, but a retry storm can consume browser capacity or API credits; use bounded retries with backoff and avoid retrying permanent validation failures.
In a self-hosted deployment, browser startup, memory use, concurrent navigation, and process cleanup become application concerns. Queue jobs if bursts could exceed safe concurrency, enforce a maximum render duration, and monitor failed navigations separately from successful captures. With a hosted API, inspect the provider’s documented response and billing indicators so your application can distinguish an invalid request from a page that simply did not produce the expected content.
Best Value
Frequently asked questions
Can an HTML-to-image API run JavaScript?
It depends on the endpoint. The html2img getting-started guide says its raw HTML endpoint can run inline JavaScript. For other services or input routes, check the endpoint’s documentation and use a readiness wait when scripts populate the captured content.
Can I capture a page that is only available on localhost?
A hosted renderer cannot generally reach a developer’s local machine through a URL such as http://localhost. Use a self-hosted browser process with access to that address, or expose an appropriately secured staging page reachable by the rendering service.
Is a full-page screenshot equivalent to a PDF?
No. A full-page screenshot is a raster image of a page; a PDF can paginate content and has document-oriented sizing and print controls. Choose based on how the recipient will view, print, or process the result.
Which approach should I start with?
Start with a hosted API if you want to generate captures without managing browser infrastructure. Start with Playwright or Puppeteer if you need browser-level behavior or local execution and are prepared to operate it. Test the hardest representative page before committing to either workflow.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
