The price tracker had run without complaint for weeks, then a Tuesday batch came back with empty rows where the product grids should have been. The saved page titles told the story: a challenge screen asking the visitor to wait, served to a script that kept retrying from the same server address.
That empty row is an anti-bot wall: the site decided the visitor was a script and served a challenge instead of data. Walls check different things (how the browser looks, where traffic comes from, whether a puzzle gets solved, whether the visitor is signed in), and Solari answers each with a setting on the same cloud browser session, so the extraction code stays put. This guide matches each wall to its setting, then covers what an escalated run costs.
Start Every Site on the Default Session
The cheapest browser that works is the right first try, so a scraping job starts on the default session and escalates only where a site objects. Solari's default mode is the fastest to start and to load pages, and it suits public pages, search and structured-data scraping. Switch to stealth the moment blocks or challenge pages appear.
Spotting that moment takes evidence, not guesses. A session created with the recording flag can be replayed in the console or downloaded as a line-by-line log, so the challenge page is on screen rather than inferred from a missing selector. Recording captures input values by default, so a session that types a password should run unrecorded. Once a replay shows the wall, the setting follows from it.
Match Each Wall to a Setting
Each wall checks a different signal, so each has its own setting, and the figure below pairs them.
The first row is the most common. Sites with bot defenses read a browser fingerprint, the mix of version quirks, fonts and rendering details that separates a script from a person. Stealth mode makes the Solari browser look like an ordinary person's; it is slightly slower and bills at a higher per-second rate, which is why it comes second.
The next two rows are about the network. A proxy server sits between the browser and the site, so the site sees the proxy's address instead of a data-centre one. Solari proxies come as rotating residential IPs by default, static ISP for a fixed IP all session, or mobile carrier IPs for the strictest targets. Sessions that share a proxy.session label share an exit IP, so a multi-step flow does not trip a check when the address changes.
Proxies require stealth, so the network fix sits on top of the fingerprint fix. The escalated session is a single create call, with the extraction code unchanged:
The last two rows differ in kind, because they ask the visitor to prove something.
Getting Past Puzzles and Login Walls
Puzzles and login walls test the visitor rather than the browser, so their settings act on the page itself. With captcha solving on a stealth session, Solari solves a CAPTCHA from reCaptcha, hCaptcha or Turnstile in the background as soon as the widget appears. Solving is not instant, so the agent waits for the widget's response field to fill before it submits.
A login wall is better cleared once than on every run. A Solari profile saves a login with its cookies and site data, and every session that attaches it by profileId starts signed in. A person can sign in once through the profile editor in the console, the easy route past two-factor prompts, and the profile then deserves the care of a password.
Each of these settings adds a line to the bill, which is the next thing to plan for.
What an Escalated Run Costs
Escalation costs money in three places: the browser hour, the proxy traffic and the solved puzzles. Stealth, proxies and captcha solving need a paid plan, and Solari bills browsers at $0.07 per hour, captcha solving at $0.005 per solve and proxy traffic at $4.00 per GB on the Professional plan, against $0.01 per solve and $6.00 per GB on Starter.
Proxy bandwidth is usually the line that grows. The blockHeavy option in the snippet above refuses video, audio and oversized images before their bodies download, so they never cross the proxy.
Width is the other limit. Stealth sessions sit under a separate stealth cap counted on top of the plan's browser limit, and a stealth session over it gets a 429 even when browser slots are free. So the design splits in two: a worker pool sized to the cap for walled sites, and default sessions for everything else.
Choosing the Right Setup
A scraping agent that meets anti-bot walls needs a ladder, not one heavy configuration. Start every site on a default session with recording on, read the replay when rows come back empty, and add stealth, a proxy tier, captcha solving or a profile only where a site asks for them. The Solari introduction shows the sandboxes on the same account, which suit cleaning the scraped data once it lands.
Get Started With Solari
The first run to try is the tracker's failing site: a default session with recording on, then the same site on stealth with a proxy country. Solari browsers are ready in 8 ms and Solari is about 2x cheaper than Browserbase per browser hour, so a fleet that escalates only where it must stays fast and cheap. Create a free Solari account to build on default sessions, then move to a paid plan when the first wall appears.
