Web Access API

Access any page.
Even the hardest ones.

One HTTP call against any URL. String returns clean content, and you only pay when it comes back.

Get an API key ↗Read the docs

First request in seconds · Free to start · Pay only on success

Live · fetch

$ curl https://request.usestring.ai/v1/fetch \

-H "Authorization: Bearer sk_live_…" \

-d '{ "url": "https://www.zillow.com/homedetails/…" }'

200 · 412ms · 18.4 KB clean text

<!doctype html> … fully rendered, anti-bot cleared, no proxy to manage.

Fast

0.41s

p50 latency

0.9s

p95 latency

Warm browser pools and a rendering path that skips a cold spin-up wherever a target allows it. Concurrency holds the tail flat under load, so a burst of requests doesn't turn into a queue.

Reliable

Reliable means it gets through the systems that others can't, and recovers on its own when a target shifts.

CloudflareDataDomePerimeterXAkamaiKasadaCAPTCHA challenges

Every request escalates automatically: lightweight request-based fetches, full browser rendering, premium proxies, geolocated routing, and CAPTCHA solving - which compliance-sensitive teams can switch off per request.

Who it's for

AI application companies

Provide your agents access to live web data better, faster and cheaper than Firecrawl, Browserbase, Exa, and Parallel.

Want to give your users the ability to query live web data? Plug in our API / MCP and see their satisfaction rates jump.

GTM teams

Enrichment: Build agents that can use String to pull live web data for any deal or contact to superpower your sales teams.

Every GTM engineer's dream: a Claygent that actually works, that doesn't get blocked, and that's way more affordable.

Data companies

Quality: Access data you previously couldn't due to anti-bot or engineering-capacity limits - we get you data from the whole web.

Tired of customers finding issues with your data - stale, lacking coverage? Use String's API and get access to the entire web without issues. Customers have reported a 50%+ reduction in triage tickets.

Vibe coders

Ship without a scraping stack: point a script at any page and get clean data back, no proxies, retries, or bot walls to babysit.

Building a side project or an agent on a weekend? Pull the pages you need with a single call instead of losing Saturday to rotating proxies and a Cloudflare challenge that wasn't there yesterday.

Q3 2026 · re-run monthly

The Web Data Frontier Benchmark, in full.

Every provider, every target, same time, head to head. The two numbers that decide whether a scraper is worth running: did it get through, and how fast.

100 hard targetsAll 16 tested providers500 requests per providerdata generated September 16, 2026View the open-source benchmark on GitHub ↗

The full benchmark, always current →Why we built our own benchmark →

Benchmark methodology

Real pages, real defenses, verified content.

We tested String Web Access API against fifteen other web access APIs by loading 100 real-world pages known to be difficult to scrape, including targets protected by anti-bot and WAF systems such as DataDome, Akamai, Cloudflare, and Kasada.

The suite covers pages with high scraping value: product detail pages, search results, social platforms, and similarly protected surfaces. Each URL was requested 5 times per provider - 500 requests each - and a request only counted as successful when the API returned a 2xx and the response contained the page's expected content.

Latency includes every target. Content uses the 75th percentile of successful attempt latencies. A complete miss uses the 75th percentile of providers that returned content for that target, or the 90-second timeout when nobody succeeds.

01 · Overall success rate

All targets, all requests. Higher is better.

string

97.0%

scrapfly

86.2%

scraperapi

84.0%

firecrawl

80.2%

apify

77.4%

bright

74.6%

scrapingbee

73.0%

context_dev

72.0%

oxylabs

69.0%

nimble

68.6%

zyte

68.0%

decodo

50.6%

scrapingdog

45.6%

browserbase

41.4%

zenrows

41.2%

scrapingant

36.4%

0%25%50%75%100%

02 · Latency score

Mean per-target p75 latency; complete misses use that target's successful-provider p75. Lower is better - read it alongside success rate.

string

7.06s

scrapfly

19.79s

scraperapi

12.97s

firecrawl

9.11s

apify

20.64s

bright

15.62s

scrapingbee

21.96s

context_dev

16.32s

oxylabs

15.86s

nimble

21.54s

zyte

17.46s

decodo

25.41s

scrapingdog

26.55s

browserbase

17.73s

zenrows

18.80s

scrapingant

18.95s

0s9s18s27s36s

Get an API key ↗Free to start · Pay only on success

Data generated September 16, 2026 (Q3 2026 run).

One call. That's the integration.

Send a URL, get clean content back. No proxy list to rotate, no headless browser to babysit.

Request · curl

curl https://request.usestring.ai/v1/fetch \
  -H "Authorization: Bearer sk_live_…" \
  -H "Content-Type: application/json" \
  -d '{ "url": "https://hard-target.com", "format": "markdown" }'

200 · 412ms · 18.4 KB clean text

Give your agents the same reach.

Add String over MCP and your agent fetches any page, even the hardest ones, as a single tool call, in Claude, ChatGPT, Cursor, Perplexity, or anything that speaks the protocol.

Already on Apify? The same engine ships as 30 ready-to-run Actors on the Apify Store. See the list.

MCP · Claude / ChatGPT / Cursor / Perplexity

{
    "mcpServers": {
        "string": {
            "url": "https://mcp.usestring.ai/v1/mcp",
            "headers": { "Authorization": "Bearer sk_live_…" }
        }
    }
}

Frequently asked questions

What is the Web Access API?

A web scraping API for humans and agents. One HTTP call against any URL returns the page as clean content, fully rendered and past whatever anti-bot is sitting in front of it. You only pay when it comes back.

Which anti-bot systems does it get through?

Cloudflare, DataDome, PerimeterX, Akamai, Kasada, and CAPTCHA challenges. Every request escalates on its own: a lightweight fetch first, then full browser rendering, premium proxies, and geolocated routing if the target needs it. CAPTCHA solving is part of the same escalation, and teams with compliance constraints can turn it off per request.

How fast is it?

0.41s at p50, 0.9s at p95. Browser pools stay warm and the rendering path skips a cold spin-up wherever a target allows it, so a burst of requests doesn't turn into a queue.

How reliable is it?

String passed 97.0% of requests in the open-source Web Data Frontier Benchmark, the best success rate of any provider in the run. The run covers 100 hard targets at 500 requests per provider, and repeats monthly.

How many requests can I run at once?

Up to 10,000 fetches and 5,000 browser sessions at once, on every plan. The request rate is what grows with your plan, and higher rates are available on request.

Can my agent use it?

Yes. Add String over MCP and your agent fetches any page as a single tool call, in Claude, ChatGPT, Cursor, Perplexity, or anything that speaks the protocol.

How much does it cost?

Start with 20,000 free standard fetches, no card needed. After that, plans start at $19/month, and the fee comes back as usage credit billed per request. Blocked requests cost nothing.

Get your API key now.

Get an API key ↗