Solari for GTM Agents: Prospect Research, Enrichment and Outreach at Scale

·

5 min read

·

Summarize with AI
ON THIS PAGE

Solari runs GTM agents at scale from one account and one API: a cloud browser per agent that is created in 8 ms at the median on Steel's open-source benchmark across 5,000 runs, saved logins in profiles, stealth for the sites that block automation, and recording for every session that touches outreach. The agent, the model and the prompts stay with the team that built them; Solari runs the environment the agent works in.

Your GTM agent works on ten accounts and then breaks on the full list. Sessions crawl, the logins it needs have expired, and the sites you care about most show it a challenge page instead of data. Each of those walls is an infrastructure problem, not a prompt problem, and each has a Solari setting that fixes it.

What a GTM Agent Needs From Its Infrastructure

Solari covers the four things a GTM agent needs: fast parallel browser sessions, persistent logins, anti-bot handling for the sites that need it, and an audit trail for anything it sends. A fifth, a machine for the agent's own scoring code, comes from the same API once the list gets long.

GTM job

What the agent needs

Solari feature

Guide

Prospect research

Many short browser sessions in parallel

Cloud Browsers

Sessions

Enrichment behind a login

A saved login every session reopens

Profiles

Profiles

Sites with bot defenses

Stealth, proxies and captcha solving

Stealth mode

Stealth

Outreach

A record of what was typed and sent

Session recording

Recording

Scoring and list cleanup

A machine for the agent's own code

AI Sandboxes

Sandboxes

Prospect Research: Session Start Is Paid on Every Account

Solari averaged 199.9 ms end to end on Steel's open-source benchmark, which times create, connect, first page load and release over 5,000 runs. The medians were 8 ms to create, 24 ms to connect, 155 ms to load the page and 2 ms to release. Research is many short sessions, so that start time is paid once per account and adds up across a territory list. Solari's results sit in PR #4, which is open, and anyone can rerun the benchmark on their own account.

Against the averages in the browserbench README, Solari ran about 4x faster than Kernel, about 4.5x faster than Steel and about 14.8x faster than Browserbase end to end.

Solari runs 150 concurrent browsers on the Professional plan for $200 a month and up to 100,000 concurrent browsers on Enterprise (pricing).

Moving an existing research agent to Solari is a one-line change. Solari sessions accept any recent CDP-compatible client through cdpEndpoint, so playwright-core, puppeteer-core and browser-use connect without a rewrite.

Enrichment Behind a Login: Profiles

Solari keeps an enrichment agent signed in between runs with profiles, so the agent reopens a saved login instead of signing in every time. A profile saves the login, the cookies and the site data. You create a profile, sign in once, and pass its profileId when a session starts.

For sites with two-factor prompts or captchas, open the profile in the console editor and sign in by hand in a live browser. If your agent already has a Playwright storageState file, upload it to the profile instead.

One detail is worth knowing up front: a profile seeds the session's default browser context. To carry the login into a context your agent creates later, apply the saved state to that context too.

Solari's Professional plan includes 500 profiles (pricing). Solari's own team runs this in production: Pinetree Health's Canva scheduling stays signed in through a Solari profile. Treat every profile like a password, because anyone with the API key can attach it to a session and act as that account.

Sites That Block Automation: Stealth, Proxies and Captcha

Solari turns on stealth per session, so the agent uses it on sites with bot defenses and runs the fast default browser everywhere else. Stealth makes the browser look like an ordinary person's, and proxies and captcha solving build on it on every paid plan from Starter up.

On Professional, Solari bills proxies at $4.00 per GB and captcha solving at $0.005 per solve (pricing). Solari runs 50 concurrent stealth sessions on Professional and 75 on Team and Enterprise, counted alongside the browser cap. For a GTM fleet that number matters more than the headline browser count, so size the protected part of the fleet against it, and keep the agent within each site's terms.

const browser = await client.launch({
  stealth: true,
  captcha: true,
  recording: true,
  profileId: "prof_abc123",
  proxy: { country: "us", session: "warmup-1", sessionDuration: 10 },
})
const browser = await client.launch({
  stealth: true,
  captcha: true,
  recording: true,
  profileId: "prof_abc123",
  proxy: { country: "us", session: "warmup-1", sessionDuration: 10 },
})
const browser = await client.launch({
  stealth: true,
  captcha: true,
  recording: true,
  profileId: "prof_abc123",
  proxy: { country: "us", session: "warmup-1", sessionDuration: 10 },
})

Outreach With a Person in the Loop

Solari records any session on request, so the agent can research and draft while a person approves each message before it goes out. Create the session with recording on, then open a temporary replay link or download a line-by-line log of everything that happened on the page. Recording captures input values by default, so sign in through the profile editor rather than recording a session that types a password.

Solari keeps replays for 30 days on Professional and 90 days on Team. Solari's own marketing system works this way: it runs in a Solari sandbox, and every outbound post waits in an approval queue for a person. The sandbox is also where list scoring and deduplication code can run, and a paused sandbox keeps its memory and processes.

What It Costs and How Solari Compares

Solari bills browsers at $0.07 per hour on the Professional plan and $0.05 on the Team plan (pricing); a GTM fleet adds only proxy traffic and captcha solves on its stealth sessions.

Browserbase is the closest peer for this work: Browserbase says its Startup plan is $99 a month with 100 concurrent browsers and 500 browser hours, then $0.10 per browser hour, on its pricing page, so both Solari rates sit below that overage rate. Browserbase is a good fit for teams that want its search and fetch APIs on the same bill. Apify and Bright Data fit teams that want ready-made scrapers or data products rather than an environment for their own agent. Browserbase is the more established vendor with the larger ecosystem; Solari is the younger company and moves faster on speed and price.

Where to Start With a GTM Agent

Solari is built for teams writing their own GTM agents: start with one research job, add a profile for the first site that needs a login, and turn on stealth only where challenge pages appear. The introduction to Solari covers the rest of the platform. 100+ companies and 2,000+ engineers already build on Solari, and you can start on Solari's free plan today, open your first session from the sessions guide and scale the same code to a full fleet.

FAQ QUESTIONS:

  1. How do I keep a GTM agent signed in to a CRM or data tool between runs?

  2. Do prospecting agents need stealth mode, proxies and captcha solving?

  3. How many browser agents can I run in parallel on Solari?

Share this article

Give every agent
a place to work

Browsers, sandboxes and desktops from one API. Start free and scale when you need to.