News

Your GPU stops after a stretch of inactivity, so you are never billed for a machine you forgot about. That timeout used to be the same for everyone. Now it is yours to set.

• Set LLM and Media timeouts independently on the Settings page.
• Choose anything from 5 minutes up — or switch auto-stop off entirely for long jobs.
• Media treats a busy ComfyUI queue as activity, so a render that runs for hours keeps the pod alive. The countdown only starts once the queue is empty.
• LLM counts your own session, so your timeout never cuts anyone else off.

Running out of balance always stops a GPU, whatever you choose here.

You can now see exactly how much GPU time you have used and what it cost — right on the Monitor page.

• Every LLM and media (image / video / voice) session is listed with its length and price.
• A totals line shows your lifetime spend, split between LLM and media.

Open Monitor from the top menu to review your usage anytime.

You can now trigger ready-made OSINT lookups directly in chat — no need to know which tools to run or in what order.

Triggers:

• osint:email <address> — checks which platforms an email is registered on, plus domain/MX lookup
• osint:person <name> — public profile and username enumeration across platforms
• osint:company <name> — resolves a company name to its official domain, then maps infrastructure and org info
• osint:domain <domain> — whois, subdomain enumeration, DNS records
• Send osint on its own to see the full list again

Each workflow runs real tools (whois, dig, holehe, sherlock, theHarvester, crt.sh) on your own rented Kali Linux VPS — rent one in the Virtual Machines section if you don't have one yet. If you don't have a machine, the model will tell you so directly instead of guessing or making anything up.

This page — the one you're reading right now. Short announcements about new features and changes, in English, Russian, and Chinese.

Nothing complicated: new posts show up here as they're published, newest first.