How to Handle CAPTCHA While Web Scraping with a Fingerprint Browser

TL;DR

• Handle CAPTCHA in a permitted scraping workflow as an explicit verification stage, and verify the destination page before extracting records.
• Nstbrowser provides persistent browser Profiles and a built-in solver for supported CAPTCHA categories. Neither capability grants access to restricted data or guarantees that a site will accept automated traffic.
• Configure CAPTCHA processing in a Profile or Group, connect the Agent, and restart the browser environment after saving changes.
• Applied means a result was applied to the page. A scraper still needs to confirm the expected application content, response, or navigation before continuing.
• Reduce unnecessary requests through authorized APIs, caching, checkpoints, and compliance with site-defined limits. An HTTP 429 response calls for slowing down and following any Retry-After instruction.
• Successful solver returns are billable, including returned results later rejected by the site. Limit retries and track CAPTCHA charges separately from collected-record counts.

Why a scraping workflow encounters CAPTCHA

A scraping workflow encounters CAPTCHA when a website requires verification before allowing a page or action to proceed.

The decision is site-specific. A page may show a visible puzzle, present a Turnstile widget, or use score-based verification without a conventional image challenge. Google’s reCAPTCHA version guidance explains that v3 returns a score rather than requiring the same interaction as a v2 checkbox challenge.

If you are researching how to bypass CAPTCHA while web scraping, define the permitted outcome first. This guide covers handling supported challenges within an authorized browser workflow. For restricted resources, obtain approved access or use the site’s available API or export route.

Repeated requests are not a reliable way to make a verification requirement disappear. They can also make the collection job harder to reconcile: several attempts may correspond to the same record, while a returned CAPTCHA result still has its own cost.

Choose an authorized collection route before configuring a browser

Use the site’s approved API, feed, or export if it meets your collection requirements, and use browser automation when the permitted workflow actually requires a browser.

Before building the job, identify the allowed resources, permitted request rate, authentication requirements, and the fields you need. Respect the site’s robots rules and terms. The Robots Exclusion Protocol specification explicitly distinguishes robots rules from access authorization; an allowed robots path is not permission to collect otherwise restricted material.

Reduce repeated work through caching and checkpoints. If the site responds with HTTP 429 Too Many Requests, pause and follow its Retry-After instruction when provided. Keep limits aligned with the site’s rules rather than changing browser identities to evade a limit.

These choices improve the design of the job without promising that CAPTCHA will never appear. Challenge frequency and acceptance remain decisions made by the destination service.

Use a persistent Profile for the intended session

A persistent Nstbrowser Profile keeps the browser environment associated with the collection session across launches.

Nstbrowser’s Profile information distinguishes persistent Standard Profiles from temporary Once Profiles. A Standard Profile retains cookies, login sessions, fingerprint settings, and local data. A Once Profile clears its session data when it ends.

For an authorized collection task that depends on an ongoing session, start with a persistent Profile assigned to the correct team. Keep its account access and collection purpose documented. Preserve the relationship between that Profile and its checkpoint so that a restart does not cause your job to treat an interrupted request as a new completed record.

Browser fingerprinting for web scraping concerns the browser properties visible to a destination website. Nstbrowser’s fingerprint-browser product supports configurable environments. Configuration does not establish permission, eliminate website controls, or guarantee that a particular target will accept the session.

Configure CAPTCHA handling before launching the job

Enable the supported CAPTCHA categories in the Dashboard before starting the browser environment used by your collection workflow.

Use these English-translated menu names to find the corresponding entries in your interface language:

1. Select the team that owns the Profile and check its Credits under Billing → Current Period.
2. Create or edit the Profile and open Advanced → CAPTCHA.
3. Enable CAPTCHA solving and select Custom configuration, or choose Follow Group settings for a shared policy.
4. Select the required CAPTCHA types in the active Profile or Group configuration and save.
5. Confirm that the Agent is connected. Stop an already running environment, then start it again.
6. Open the Nstbrowser plugin and review its status before navigating through the intended workflow.

The supported categories are reCAPTCHA v2, reCAPTCHA v3, Text CAPTCHA, supported AWS CAPTCHA image challenges, and Cloudflare Turnstile. Standard and Enterprise reCAPTCHA variants share their respective v2 or v3 selection but have different usage rates.

A Group policy helps when several authorized collection Profiles need the same selection. A custom Profile configuration takes precedence over inherited Group settings. Restart affected environments after either kind of saved change.

Add a verification checkpoint before extraction

Put a verification checkpoint between navigation and extraction so your workflow does not mistake a challenge page for the requested data page.

An HTTP response or a loaded browser tab does not, by itself, establish that your expected business content is available. Design a bounded workflow around observable page outcomes:

For example, an owner-authorized catalog collection might require a known product identifier and price element before accepting a page as a product record. If the page instead contains a verification form, retain the checkpoint as pending. Do not export the verification message as product text.

This is a workflow design example. The selectors, required fields, and success markers must come from the actual application you are authorized to collect from.

Read CAPTCHA status and resume the correct task

Read the solver state to understand the verification stage, then use the website’s own result to decide whether extraction can resume.

For manual observation, the built-in plugin displays Checking, Allowed, Solving, Applied, and Disabled. Wait during Solving; use the Dashboard and an environment restart to change disabled or incorrect configuration.

Applied means the result has been applied to the page. A form submission or server-side check may still be required. Continue only when the page meets your application-level success condition.

For developer-controlled automation, Nstbrowser also documents a separate page-level status interface. Its machine-readable states differ from the plugin’s display labels. Use the interface documented for your script and keep failure handling separate from record extraction.

If a page does not require CAPTCHA, wait for its normal content condition. Requiring a CAPTCHA terminal state on every URL can stall pages that never create a CAPTCHA task.

Bound retries and track the right errors

Use a finite retry policy with a clear stop condition so that an unresolved verification step does not become an endless request loop.

Separate navigation failures, missing content, CAPTCHA-processing failures, and website rejection in your run log. A retry should address a known recoverable condition, such as adding funds to the correct team, rather than repeatedly submitting the same unresolved page.

After a failed attempt, record enough context to resume safely: the Profile identifier, job checkpoint, page category, relevant state, and attempt count. Keep credentials, cookies, personal page content, and verification material out of general logs. Collect only the fields needed for the authorized task and apply your team’s retention policy.

For Turnstile on an application you own, Cloudflare recommends official test keys for automated tests. Its production browser-support policy also excludes automated challenge solving. A controlled test validates your flow; it does not promise production challenge acceptance.

Budget for solver returns rather than extracted rows

CAPTCHA usage charges follow successful solver returns, while extracted rows describe the output of your collection job.

Nstbrowser deducts CAPTCHA fees from the Credits of the team owning the Profile. A task that fails before returning a result is not charged. A result successfully returned by the solver remains billable if the website later rejects it, and a successful retry is charged at the applicable rate.

For a simple cost illustration, 3,000 successful reCAPTCHA v2 returns at the October 19, 2026 Standard Growth rate of 1.20 per 1,000 cost **3.60** in CAPTCHA fees. This does not imply 3,000 accepted pages or extracted records, and excludes browser, subscription, and other service costs.

Check the plan’s current rates before budgeting. Review counts, unit prices, and amounts under Billing → Usage Analysis → CAPTCHA, then review Credit transactions separately. There is no fixed conversion between the number of browser Profiles, the number of collected rows, and solver usage.

Continue collection from a verified page

Resume the collection job only after the intended page content is available and its required fields pass validation.

Keep the sequence explicit: permitted access, correct Profile, supported CAPTCHA handling when needed, website acceptance, extraction, and a saved checkpoint. If a challenge remains unresolved or access is not permitted, stop that path and use the site’s native verification, approved API, or support route.

Nstbrowser brings Profile management and supported CAPTCHA handling into the same browser workflow. The trial button opens the product; solver and other usage charges remain separate.

FAQ

Q: Can I bypass CAPTCHA while web scraping?

This workflow handles supported CAPTCHA challenges within an authorized session. It does not remove a website’s access requirements or guarantee acceptance of automated traffic. Obtain permission and use approved data-access methods for restricted resources.

Q: How can I avoid unnecessary CAPTCHA interruptions when web scraping?

Use an approved API or export where suitable, avoid redundant requests through caching and checkpoints, and follow site-defined request limits. These practices reduce unnecessary collection work; they do not guarantee that a site will stop requesting verification.

Q: Does a fingerprint browser solve every CAPTCHA automatically?

No. Browser configuration and supported CAPTCHA processing have defined scopes. Nstbrowser documents five configuration categories, while the website retains control over final acceptance and any additional access checks.

Q: How do I know when extraction can continue?

Check the website’s expected content or application-level success condition before extracting. Applied means that a result was applied to the page, not that the requested data is ready or the business action completed.

Q: Should every page wait for a CAPTCHA status?

No. A page without a CAPTCHA may never produce a CAPTCHA completion state. Wait for the normal content condition on those pages, and add CAPTCHA-specific handling only where the workflow actually requires it.

Q: Why is my solver bill higher than my successful-record count suggests?

Billing follows successful solver returns. Returned results can be charged even when a website rejects them or extraction later fails, and successful retries are also billable.

Q: Can I keep retrying until the page accepts a result?

Use a bounded retry policy and stop when the permitted workflow cannot proceed. Unlimited retries can repeat unresolved failures and create additional billable returns without producing valid records.

Q: Does robots.txt establish permission to scrape a resource?

No. Robots rules communicate crawler preferences rather than access authorization. Review the site’s terms, permissions, and applicable data-handling requirements in addition to respecting its robots rules.