Matthew Berman showed 11 AI tools in this video on . 6 are verified available in the EU. Each row below jumps to the exact second. Part of week 2026-40.
Tools shown in this video, with timestamps
OpenAI's always-on, proactive personal AI assistant inspired by OpenClaw, running in the cloud with its own environment and connecting to Gmail, Docs, Calendar, Slack and Teams. It lives inside the ChatGPT app and controls ChatGPT and Codex threads; chats with Dots don't count toward usage, but controlling threads does. The creator likes the idea but would prefer a standalone app.
GPT-6 Astra running on Cerebras chips, eight times faster than on Nvidia GPUs but about six times more expensive. The creator burned $1,000 in tokens in an hour and a half and showed a side-by-side demo where Ultrafast launched a rocket before standard finished thinking.
Live demoNew $500/month ChatGPT plan with 25x the usage of Plus. The $200 plan drops from 20x to 10x, so the creator says users get barely more than before at the old $200 tier for much more money.
A smaller, cheaper, faster model that is nearly as good as GPT-6 Astra, priced at $2/M input, $10/M output and $0.10/M cached input. Benchmarks shown include a developer-preference benchmark, GDPval-style PDF tasks and OSWorld, where it matches or is near Astra at a fraction of the price.
OpenAI's frontier large model, priced at $10/M input and $50/M output, used as the comparison point for Sol. It performs slightly better at the top end of thinking tiers on OSWorld.
AI that continuously scans your code for issues and vulnerabilities and reports them.
Codex now runs natively in the cloud so the laptop doesn't need to stay open. The creator says local tool calls and terminal commands are now the bottleneck with Ultrafast mode.
Refreshed Codex CLI that can be controlled via voice, along with a new code review experience in Codex.
Preview API of a model built for quick decisions, based on Luna, OpenAI's lightest and smallest model. The creator thinks a purpose-built decision model like Jev would still be faster.
Relaunch of apps/plugins natively in ChatGPT, plus Sign in with ChatGPT for third parties, letting users bring their own ChatGPT tokens instead of paying separately. The creator calls bringing your tokens generous.
A Notion-like collaborative workspace native to ChatGPT where teammates and agents work together on presentations, spreadsheets and websites.
