Launching November 10, 2026
String

String

Web scraping API that accesses any page, even the hardest ones.

SubscriptionAPIAI Agents5 impressionslaunches Nov 101 comment

Comments

>log in to comment
  • Josh Parmar[maker]· 22h ago

    Unlike typical scrapers that need proxies and custom retries, String lets you fetch any page—including hard, bot‑protected sites—with a single HTTP call and only pays on success. It returns fully rendered, clean text in under a second, and scales without queuing. I built it to simplify live web data for AI agents and GTM tools—what do you think?

Web scraping API that accesses any page, even the hardest ones. One HTTP call against any URL - String returns clean content, and you only pay when it comes back. One HTTP call against any URL. String returns clean content, and you only pay when it comes back. First request in seconds · Free to start · Pay only on success String fetches more pages than any other provider17 providers · 100 bot-protected sites · String98.6%View the benchmark→ 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 means it gets through the systems that others can't, and recovers on its own when a target shifts.
  • Web access for everyone
  • Integrates with any workflow or platform
  • Handles anti‑bot, CAPTCHA, proxy rotation automatically
  • MCP server lets agents skip the HTTP layer

String provides a web-scraping API that returns clean, fully rendered page content with a single HTTP call.

for
AI app developers, GTM engineers, data providers, and solo developers needing reliable scraping.
pricing
freemium · free trial
license
MIT
usestring/string-ai-mcp 5 2TypeScriptupdated a day ago
works withcURLPythonNode

Key features

6 features of String
  • One-call fetch — Send a URL and receive clean, rendered content without managing proxies or browsers.
  • Warm browser pools — Keeps browsers warm to avoid cold spin-up latency, achieving 0.41s p50 latency.
  • Anti-bot bypass — Handles Cloudflare, DataDome, PerimeterX, Akamai, Kasada and CAPTCHA challenges automatically.
  • Scalable concurrency — Supports up to 10,000 fetches and 5,000 browser sessions concurrently on every plan.
  • Pay-only-on-success — Only billed when the request returns a successful response; blocked requests are free.
  • MCP integration — Can be added to AI agents via MCP for use in Claude, ChatGPT, Cursor, Perplexity, etc.

Use cases

  • Enrich AI agents with live web data without building a scraping stack.
  • Provide sales teams with up-to-date web data for leads and deals.
  • Collect data from bot-protected sites for data-as-a-service platforms.
  • Build side-project tools that need occasional page fetches without proxy management.

String pricing

  • Free$0≈20k fetches free · No card required · Pay only on success
  • Starter$19/month≈76k fetches/month · Standard fetch $0.25/1k · Premium fetch $2.40/1k
  • Growth$99/month≈619k fetches/month · Standard fetch $0.16/1k · Premium fetch $0.90/1k
  • Scale$499/month≈3.8M fetches/month · Dedicated Slack support · Higher request rate
  • Volume$999/month≈9.1M fetches/month · Priority Slack support · Lowest rates

String vs alternatives

StringPostmanLlamaIndexVonageReplicate
Best forSingle-call web scraping for hard-to-reach pagesAPI development and testingLLM data indexingSMS/voice/video APIsRunning/fine-tuning AI models
PricingFreemiumSubscriptionFreeSubscriptionSubscription
DevHunt upvotes10000
LaunchedNov 2026————

String FAQ

What does the Web Access API return?+

It returns the full page content, fully rendered and cleaned, in a single HTTP response.

Which anti-bot systems can it bypass?+

It works through Cloudflare, DataDome, PerimeterX, Akamai, Kasada and CAPTCHA challenges, escalating automatically as needed.

How fast is the service?+

Typical latency is 0.41 s at p50 and 0.9 s at p95, with warm browser pools preventing queue buildup.

What is the success rate?+

String achieved a 98.6 % success rate on the Web Data Frontier Benchmark, the highest among tested providers.

What concurrency limits apply?+

All plans allow up to 10,000 concurrent fetches and 5,000 concurrent browser sessions; request rate varies by plan.

How is pricing calculated?+

You receive a monthly credit (e.g., $19 for Starter) and each successful request draws down that balance at per-1,000-request rates; blocked requests are free.

Summarized by DevHunt from usestring.ai · Nov 2, 2026. Details may change; check the official site.

Around the web

Articles, videos and discussions