ushakov

thank you so much, great to hear the thinking behind these considerations :)

> We don't spin up a full browser instance for every request (which is slow and expensive)

there's really no excuse for not spinning up a browser every request. a Firecracker VM boots ~50ms nowadays

> We respect robots.txt rules.

you might, but most companies in the market for your service don't want this

what does Katakana add on top of Kata?

hey, I work at E2B, anything we can do to improve the setup for you?

the browser looks like it is based on Chromium indeed

thanks for explaining, especially the runtimes part!

i am currently running Docker MCP Containers + MCP Gateway mixed with Remote MCPs in microVMs (aka. Sandboxes).

seems to be the most portable setup, so you don't have to worry about dealing with different exec like uvx, poetry, bun, npx and the whole stdio/streamable http conversion.

lambdas sound interesting, esp. if you have figured out the way to make stateful work stateless, but comes with the downside that you have to maintain all the integrations yourself + the environment itself might have compatibility issues. i've seen someone also using cloudflare dynamic workers for similar use-case (disco.dev), but they're maintaining all the integrations by hand (or with Claude Code rather). more extreme version of this would be writing custom integration specific to the user by following very strict prompt.

anyways, i'll look into Metorial as am curious about how the portable runtimes work.

i am also maintaining a list of MCP gateways, just added you there as well: https://github.com/e2b-dev/awesome-mcp-gateways

thanks for building this, looking forward to checking it out!

congrats on the launch!

why do I need a specialized platform to deploy MCP instead of just hosting on existing PaaS (Vercel, Railway, Render)?

also if you're not using VMs, how do you isolate per-user servers?

Looks awesome :)

We're doing something similar at E2B, we should chat!

official source: https://www.bart.gov/news/articles/2025/news20250905

Do you think you can make it run in Jupyter as a Kernel?

i work on E2B, we are open-source sandbox runtime used by Perplexity, Manus, Hugging Face among others.

check it out: https://e2b.dev

we offer this with E2B Desktop

Demo: https://surf.e2b.dev

SDK: https://github.com/e2b-dev/desktop

hi Peter!

how long does 221(g) administrative processing take to complete in your experience? anything one can do besides waiting (Russian citizen working in tech, almost 1yr without adjustments)?

fun fact I learned about Imba is that it's name stands for "imbalance" (like in computer games!)

how is it different from Step CI?

https://stepci.com

there’s also llm-scraper: https://github.com/mishushakov/llm-scraper

disclaimer: i am the author

GCP supports nested virtualisation

everything, including the infra is open-source (below), but it currently requires more than just your laptop (gcp, nomad, firecracker, postgres, etc.)

this way, we're able to run millions secure sandbox environments

i appreciate asking though and will be forwarding to my team to see if we can come up with a way for users to emulate the execution locally

source code: https://github.com/e2b-dev/infra

have you tried e2b.dev? it runs lightweight sandboxes using firecracker, python and third-party packages

disclaimer: i work there

Happy to have met you in Berlin at the merge :)

yes, but only the text-only models!

https://www.llama.com/docs/model-cards-and-prompt-formats/ll...

Go to one of the AI Tinkerers events in your area?

https://aitinkerers.org/p/welcome

I do not understand what this actually is. Any difference between Browserbase and what you’re building?

Also, curious why your unstructured idea did not pan out?

I was talking to my friend today about dating apps being a softporn addiction

The reason your website is lagging so much is because of the globe. I have reported this a while ago: https://github.com/shuding/cobe/issues/78

There’s also llm-scraper in TypeScript

https://github.com/mishushakov/llm-scraper

In my experience Anthropic models are more steerable (requires less prompting) than OpenAI's. For example in code-generation, I'd tell GPT-4 to not include any comments, yet sometimes it would just ignore this. Have not experienced this with Opus yet.

Awesome! The problem with extracting schema automatically is that you won't know what comes out of it upfront and it could be changing on every run. What I'm trying to do is enable scraping webpages in a structured (and type-safe!) manner.

Awesome! Keep in mind there's already scrapeghost and entities-extraction-web-scraper in Python.

I've tried using it with Groq's Llama 3 70B and it worked well :)

Definitely. Smaller models like Haiku are already pretty capable (and cheap!)