Sponsored by
OpenAI's smartest model yet shipped last week with Codex baked in and agents running tasks end to end.
OpenAI released GPT-5.5 on 23 April, calling it its smartest model yet. It rolled out across ChatGPT and Codex on Plus, Pro, Business and Enterprise tiers, with a Pro variant for longer-horizon work.
The launch arrived with a system card, fresh Codex automations, workspace agents and an enterprise scaling push. The framing hints at OpenAI's longer game: a single ChatGPT super app for work and life.
What can it do?
Available in ChatGPT and Codex from 23 April across paid tiers
New GPT-5.5 Pro variant tuned for longer-horizon, higher-stakes tasks
Agentic by default: multi-step workflows with less user prompting
Deeper Codex integration: plugins, skills and scheduled automations
Workspace agents now run cloud-based jobs across connected tools
Partner message
Join 2M+ Professionals Getting Ahead on AI
Keeping up with AI shouldn't feel like a second job.
But between the new tools, viral posts, and endless hot takes, most people spend hours every week trying to figure out what actually matters.
The Rundown AI fixes that.
It's a free newsletter that gives you the AI news, tools, and tutorials you actually need to know. All in just 5 minutes a day.
Over 2M professionals at companies like Apple, Google, and NASA already read it every morning to stay ahead.
Plus, if you complete the quiz after signing up, they'll recommend the best tools, guides, and courses for your specific job and needs.
Why does it matter?
The release reframes ChatGPT from a chat surface into a work execution layer, with Codex and workspace agents handling end-to-end tasks rather than answering one-off questions.
It tightens the bundle. Pulling the model, browser ambitions and Codex inside one subscription pushes OpenAI closer to its long-rumoured super app.
Our take
A 0.5 version bump understates what just shipped. OpenAI is signalling that the next phase of competition runs through agents and integrations rather than chat interfaces and benchmark scores. Codex now has 4 million weekly active users, and Accenture, PwC and Infosys are lined up to roll it out across the software development lifecycle.
The super app framing also explains the timing. With Anthropic making serious gains so far this year, OpenAI needs a moat beyond raw model quality. Owning the workflow, from prompt to deployed code, is a more defensible position than owning the leaderboard.
And another big thing… Google's $40bn Anthropic bet
Google committed up to $40bn in cash and compute to Anthropic on 24 April, deepening a relationship that already underpins much of Claude's training infrastructure. The deal landed in the same week Anthropic's Mythos cybersecurity model triggered alarms at central banks and intelligence agencies, and as the company expanded its Amazon compute partnership by another five gigawatts.
The frontier lab race is now, in effect, a capital and compute race, with the hyperscalers picking sides and writing cheques large enough to reshape the AI map.
