@getsolari is joining a Stealth Incubator backed by @kakaocorpglobal, to host an intimate event in SF for researchers and founders building computer-use agents, across frontier labs, agent startups, and university groups
We'll dive into:
→ What stays and what dissolves as software gets rebuilt for agents
→ World models: can agents learn software environments the way they learn the physical world?
→ The Jevons paradox: what happens when computer work gets cheap
Candid, off-the-record (Chatham House Rule), plus a first look at what we're both building next
Request an invite 👉 https://t.co/wdNxrwKCnM
Seats are extremely limited
Excited to be working with @JqOnly, @dylanventures_ and @henryyam_@Techweek_
We raised a $1M pre-seed to connect agents to the world
And in the last 2 months we've launched Solari, the fastest agent infra on the market:
- 2K+ developers
- 100+ companies
- 4x faster than Kernel, 15x faster than Browserbase
- Sandboxes 1.3x faster than E2B, 1.7x faster than Daytona
Thank you to every developer and company who's tried Solari, shared feedback, and helped us build. Grateful to our investors for believing in us this early.
More to come 🚀
- https://t.co/v89Q4oCQ10
Internal tool automation is one of the most common things developers build on Solari
Legacy web dashboards, internal admin panels, old SaaS tools with no API — anything your team logs into manually and clicks through repeatedly
That is the exact thing Solari was built to replace.
The browser is the silent bottleneck killing AI agent performance
Most teams building agents today know their model is fast. They know their prompts are good. What they do not realize is that the browser underneath is adding overhead to every single task
That is what Solari is built to fix: the world's fastest agentic browser
https://t.co/eF04jlYU1V
2. Lead scraping agents
Give the agent a company name and it comes back with a full profile, from education, publications, work, resume, to the content they're posting
But rather just scraping LinkedIn, Solari enables your agent to access the ENTIRE web
3. Internal tool agents
Every team uses web-based tools with no API — legacy dashboards, admin portals, or old SaaS platforms
Solari lets agents operate them the same way your team does. If it's on the internet, Solari can do it
We bypass bot detection
4. Support agents
Automated customer service sucks, because it can't take action. Solari enables your agent to log into portals, triage tickets, update records, and close cases
No more waiting for sh*tty bots or human responses
5. Web scraping
QA automation, data extraction, browser workflows. If it runs in a browser, it runs on Solari
The fastest agentic browser 👉 https://t.co/D6t143KmD1
Our team is hiring for Growth Engineers
We're looking for someone to lead our
We’re looking for a rare kind of builder: someone who’s deeply online, plugged into X and LinkedIn, knows the AI agent landscape cold, and understands how to turn product momentum into attention.
But this isn’t just a content role. You also build amd turn a product update into something developers actually care about, all in the same afternoon
What you’ll own:
- Growth experiments across product, content, and distribution
- Turning launches, demos, and technical ideas into developer-facing content
- Building with and for our developer community
- Helping shape the roadmap based on user demand
Know someone great or think it’s you? Reply and we’ll reach out.
Introducing @getsolari Cloud Browsers
When your agent works on the internet, it needs a browser just like you do
Solari browsers are ready in 8ms, stay signed in between runs, move like a real person to avoid bot-detection, and record everything so you can replay any run when something needs fixing
Designed to be the fastest browser for agents, Solari is free to start
Check it out 👉 https://t.co/XcRxTirwnI
If you need help getting started, our team will help you get set up personally in our Discord
https://t.co/O4r9aNgC23
Solari browsers are 4x faster than #2, across 5,000 runs with 0 failures
With Solari, your agents start working the instant a task begins. Browsers open instantly, sessions run in parallel by the hundreds, and nothing waits on infrastructure
- 4x faster than Kernel (794ms)
- 4.5x faster than Steel (894ms)
- 15x faster than Browserbase (2,967ms)
- 18x faster than Hyperbrowser (3,657ms)
- 40x faster than Anchorbrowser (8,001ms)
By giving your agents faster infra, you do more in less time, at a cheaper cost
If you're building agents, start for free on Solari
👉https://t.co/hfLXDV9354
What "agent-native" actually means:
Cold start in milliseconds, not seconds
Vision-first interaction (no brittle selectors)
Persistent state across runs
Replay built in, not bolted on
Captcha solving as an API call
If your provider does not have all five, they are a browser tool with marketing.
Cold start time = how long it takes a browser to go from "I need a session" to "the session is ready"
Most developers don't think about this until they're debugging why their agent feels slow
At scale it can compound very fast. If your agent completes 1,000 tasks per day and cold start is 3 seconds instead of 500ms, you're losing 41 minutes of every single day
It's not glamorous, but it's also one of the highest-leverage optimizations in any agent stack
We're the world's fastest at 35ms
We ran a full browser benchmark across every major provider.
Solari: 199ms
Kernel: 793ms
Steel: 894ms
Browserbase: 2,966ms
Hyperbrowser: 3,657ms
AnchorBrowser: 8,001ms
Solari is 14.9x faster than Browserbase. 40x faster than AnchorBrowser. Across 5,000 runs, 100% reliability.
The bottleneck in AI right now is not intelligence. It is the ability to act.
Models are getting smarter every quarter. The infrastructure they execute on is the same it was three years ago.
print("hello world")
The final frontier of AI is execution.
Pinetree Research builds Computer-Use Agents (CUA) that use software like humans do: navigating apps, operating systems, and completing real work.
We are engineering the future.