Screenshots
PNG, JPEG, and WebP captures with full-page, viewport, scale factor, mobile, and dark mode controls.
300 free screenshots to start
Screenshot API for AI agents and developers.
Capture, read, compare, and deliver any web page through one API, official SDKs, or the SnapshotFlow MCP server — without running browser infrastructure.
No credit card for free start
Hosted API
SDK and MCP ready
Render options
Full pagetrue
FormatPNG
Viewport1440 x 900
Block adson
curl "https://api.snapshotflow.com/screenshot?url=https://stripe.com&full_page=true"
-H "X-Api-Key: sk_live..." --output stripe.png
-H "X-Api-Key: sk_live..." --output stripe.png
Status200 OK
CacheMISS
Time1.4s
Integrate your way
Official SDKs and native tools for AI agents.
Official SDK
Node.js
Typed capture, batch, diff, and webhook helpers.
Official SDK
Java
A small JVM client, without a local browser.
HTTP API
JavaScript
Use fetch from your server-side workflow.
Command line
cURL
A direct route into scripts, cron, and CI.
AI agents
Remote MCP
Connect Claude, Codex, Cursor, Copilot, or ChatGPT.
Built for AI workflows
Give agents evidence, not another browser to maintain.
A single MCP connection gives an agent the right tool for visual review, page text, comparisons, or queued captures.
Connect the MCP server- 1
Connect
Add the secure remote MCP endpoint to the client where your agent works.
- 2
Request
Ask for a screenshot, text extraction, visual diff, or a batch review.
- 3
Act on evidence
Receive an image, structured page context, change statistics, or job result.
One browser API, from first capture to production workflow
PNG · JPEG · WebP
pixel-ready images
reports and records
Markdown · HTML
content for AI
Visual diff
change metrics
Batch · Webhooks
production workflows
Puppeteer
Fastify
Sharp
Pixelmatch
Docker
MCP-ready
Input and output
Turn a URL or HTML payload into the artifact your workflow needs.
Use one API surface for screenshots, PDFs, visual diffs, metadata, and LLM-ready content extraction.
PDF export
Generate documents with page format, orientation, and background graphics for invoices, reports, and evidence.
Content extraction
Return markdown, HTML, plain text, and metadata beside a screenshot for AI agents and analysis workflows.
Visual diff
Compare two URLs pixel-by-pixel and return changed pixels, diff percentage, and a highlighted diff image.
Batch capture
Render up to 10 URLs in one request with bounded parallelism through the browser pool.
Async webhooks
Queue heavier jobs, return a job ID immediately, and notify your application when the capture finishes.
The browser work, handled
Ship the output. Skip the browser cluster.
Instead of maintaining browser processes, queues, retries, image transforms, and storage glue, send SnapshotFlow a URL or HTML payload and choose what should come back.
Your request
URL + capture options
GET /screenshot?url=example.com
&full_page=true&extract_content=true
&full_page=true&extract_content=true
01
Render
Load the real page in a browser.
02
Prepare
Apply viewport and cleanup options.
03
Deliver
Return, store, cache, or webhook.
Clean captures
Screenshots without the clutter that ruins reports.
Block ads, trackers, cookie banners, fonts, media, or your own selectors so every capture is consistent enough for audits, research, and customer-facing workflows.
Ad and tracker blockingGhostery-powered rules keep noisy pages under control.
Device emulationMobile viewport, timezone, geolocation, and dark mode.
Custom CSS and JSInject page fixes before the screenshot is taken.
Deterministic cache keysEvery output-changing option becomes part of the cache key.
Before cleanup
Raw page
AD
COOKIES
CHAT
After SnapshotFlow
Clean capture
Ads blocked
Cookies hidden
Chat closed
Developer API
Simple requests, production-shaped responses.
Paste a URL, choose the output, and let the browser pool handle the render.
Request
curl "https://api.snapshotflow.com/screenshot?url=https://stripe.com&full_page=true&format=png&width=1440" \
-H "X-Api-Key: mykey" \
--output stripe.png
-H "X-Api-Key: mykey" \
--output stripe.png
Response
200 OK - image/png
X-Cache
MISS
MISS
ETag
ss:766de460
ss:766de460
Size
350KB
350KB
Output
binary PNG
binary PNG
Request
curl -X POST "https://api.snapshotflow.com/batch" \
-H "Content-Type: application/json" \
-H "X-Api-Key: mykey" \
-d '{"urls":["https://stripe.com","https://vercel.com"],"response_type":"base64"}'
-H "Content-Type: application/json" \
-H "X-Api-Key: mykey" \
-d '{"urls":["https://stripe.com","https://vercel.com"],"response_type":"base64"}'
Response
{
"total": 2,
"results": [
{"cached": false, "image": "data:image/png;base64,..."}
]
}
"total": 2,
"results": [
{"cached": false, "image": "data:image/png;base64,..."}
]
}
Request
curl "https://api.snapshotflow.com/diff?before=https://myapp.com/v1&after=https://myapp.com/v2&response_type=json"
Response
{
"changed_pixels": 14320,
"diff_percent": 1.40,
"has_changes": true
}
"changed_pixels": 14320,
"diff_percent": 1.40,
"has_changes": true
}
Enqueue
curl "https://api.snapshotflow.com/screenshot?url=https://stripe.com&async=true&webhook_url=https://myapp.com/done" \
-H "X-Api-Key: mykey"
-H "X-Api-Key: mykey"
Webhook payload
{
"job_id": "550e8400...",
"status": "done",
"result": {"storagePath": "storage/stripe.png"}
}
"job_id": "550e8400...",
"status": "done",
"result": {"storagePath": "storage/stripe.png"}
}
Request
curl "https://api.snapshotflow.com/screenshot?url=https://stripe.com&response_type=base64&extract_content=true&content_format=markdown&metadata=true"
Response
{
"image": "data:image/png;base64,...",
"content": "# Stripe...",
"metadata": {"title": "Stripe"}
}
"image": "data:image/png;base64,...",
"content": "# Stripe...",
"metadata": {"title": "Stripe"}
}
Built around real workflows
Screenshots become more useful when they move with your work.
Capture the page, keep its context, and route the result into QA, monitoring, AI, or customer-facing products.
QA & releases
Catch visual changes before users do.
Compare two URLs pixel by pixel and return a diff image, changed-pixel count, and percentage your CI workflow can act on.
Visual regression testing
Release diff
Changes found
Changed pixels14,320
Diff1.40%
Keep a visual record of what changed.
Schedule repeat captures, use cache controls, and send completed async jobs back through webhooks.
Mon
Wed
Today
Give vision models the pixels and the page context.
Return a base64 image together with markdown, HTML, plain text, or metadata in one response.
Turn live pages into durable output.
Create full-page images or PDFs for research, audit trails, reports, and repeatable evidence workflows.
Generate consistent website visuals on demand.
Use fixed viewports and formats to create thumbnails, previews, or open graph images from HTML.
OG image generationHow it works
From request to cached artifact in a few predictable steps.
The browser pool renders the page once, applies capture options, stores the output when needed, and serves repeat requests from cache.
1
Request
URL, HTML, and render options.
2
Render
Browser pool loads and prepares the page.
3
Capture
Image, PDF, diff, or JSON payload.
4
Return
Stream, store, cache, or webhook.
Build or call
Keep control of the workflow, without owning every moving part.
A self-managed browser can take a screenshot. SnapshotFlow packages the surrounding work into the same API surface.
See the API referenceWorkflow layer
Self-managed browser
SnapshotFlow
Browser lifecycle
Launch, pool, and recycle it
Handled by the API
Output formats
Add image and PDF transforms
PNG, JPEG, WebP, and PDF
Page cleanup
Maintain selectors and request rules
Blocking plus custom CSS and JS
Large jobs
Build a queue and callbacks
Batch requests and async webhooks
Repeat requests
Design a cache layer
Deterministic cache keys
Machine-readable context
Add extraction and response shaping
Markdown, HTML, text, and metadata
Use any HTTP client. Start with cURL, JavaScript, the Node.js package, or the Java SDK.
Node.js quickstart →Free start
Start with 300 screenshots before you need anything else.
Create an account, issue an API key, and test real captures against the hosted SnapshotFlow API. When you are ready, wire it into your product or automation workflow.
Up in minutes
1
Create a free account
open "https://dashboard.snapshotflow.com/#/register"
2
Generate an API key in the dashboard
# Copy your key from Account or API Keys
SNAPSHOTFLOW_API_KEY=sk_live_...
SNAPSHOTFLOW_API_KEY=sk_live_...
3
Take the first screenshot
curl "https://api.snapshotflow.com/screenshot?url=https://example.com" -H "X-Api-Key: $SNAPSHOTFLOW_API_KEY" --output out.png
What’s new
Latest from the changelog.
New integrations, SDK releases, and practical improvements to the SnapshotFlow browser layer.
Changelog ·
New MCP Setup Guides for AI Clients
Connect more AI clients to SnapshotFlow with client-specific MCP setup and prompt examples.
Read updateChangelog ·
Screenshot API for JavaScript Is Now Available
Use fetch or XMLHttpRequest for browser screenshots, blob previews, and secure production proxies.
Read updateChangelog ·
Screenshot API for cURL Is Now Available
Capture websites from the command line as PNG, JPEG, WebP, or PDF without installing headless Chrome.
Read updateChangelog ·
SnapshotFlow MCP Server Is Now Available
Give AI agents secure access to screenshots, batches, visual diffs, text extraction, and async capture jobs.
Read updateFrom the blog
Guides, comparisons, and deep dives on screenshot automation.
Practical writing on capturing pages, rendering HTML, visual diffs, and shipping screenshots in production.
Guides ·
Cookie Banner Screenshot API: Clean Page Captures
Capture pages without consent overlays by combining cookie-banner blocking, ad blocking, selector hiding, and deterministic wait settings.
Read articleGuides ·
Batch Screenshot API: Capture Multiple URLs in One Call
Send one request for up to 10 URLs, inspect every result, and scale larger screenshot jobs in controlled chunks.
Read article
Guides ·
Website Visual Change Detection on a Schedule (2026 Guide)
Capture a page on a schedule, compare each run to the previous baseline with one /diff call, and get alerted only when the pixels actually change. Cron-ready and MCP-ready.
Read article
Article ·
Sending Screenshots to Claude and GPT: An AI Vision Pipeline
Capture any web page with one HTTP call, then let Claude or GPT analyze the pixels. Capture settings, working code for both APIs, and the MCP path for agents.
Read articleFAQ
Answers to the most common Screenshot API questions.
Quick answers on rendering, full-page captures, pricing, and AI-agent integrations.
Your first capture is the easy part
Make screenshots part of the product, not another service to maintain.
Start with 300 screenshots, explore the API with a real URL, and connect the result to the workflow you already use.
No credit card required for the free start.