OpenAI's superapp hiding inside Codex

2026-04-17 · Source: The Rundown AI · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation · Depth: Intermediate, medium

Summary

OpenAI has significantly updated its Codex platform, transforming it from a dedicated coding agent into a more comprehensive application that integrates features from ChatGPT and Atlas. This update introduces capabilities such as background computer use, allowing Codex to operate any Mac app and run multiple agents simultaneously, even in applications without APIs. A new memory preview feature enables context retention across sessions, while automations support long-running tasks. The platform also includes an Atlas-powered in-app browser for developers to direct Codex and inline gpt-image-1.5 for generating mockups. Codex has reached 3 million weekly users with 70% month-over-month growth, signaling OpenAI's strategic move towards an all-in-one "superapp" vision. Additionally, Anthropic released Claude Opus 4.7, which outperforms GPT-5.4 and Gemini 3.1 Pro in agentic coding benchmarks, though it trails Anthropic's unreleased Mythos Preview.

Key takeaway

For AI Architects evaluating platform strategies, OpenAI's Codex update signals a shift towards integrated "superapp" ecosystems, potentially simplifying complex workflows but requiring careful assessment of vendor lock-in. Your team should explore how these expanded agentic capabilities, including background computer use and persistent memory, could streamline development and operational tasks. Additionally, consider the implications of domain-specific models like GPT-Rosalind for specialized applications, as they may offer superior performance over general-purpose LLMs for targeted industry solutions.

Key insights

OpenAI is evolving Codex into a "superapp" with expanded agentic capabilities, while Anthropic's Opus 4.7 advances coding performance.

Principles

• Domain-specific models enhance performance for specialized tasks.
• Agentic capabilities are expanding beyond coding to general computer use.

Method

Install Ollama, select a lightweight model like gemma3 (3GB), download it, and then interact via chat for local LLM operation without internet.

In practice

• Use Ollama to run LLMs locally on laptops for free.
• Explore Codex's new features for multi-agent workflows and in-app browsing.
• Consider domain-specific models like GPT-Rosalind for scientific research.

Topics

• OpenAI Codex
• AI Superapp
• Anthropic Claude Opus
• Domain-Specific AI Models
• Local LLMs

Best for: CTO, VP of Engineering/Data, AI Architect, AI Engineer, Director of AI/ML, AI Product Manager

Related on AIssential

• OpenAI’s GPT-5.3-Codex drops as Anthropic upgrades Claude — AI coding wars heat up ahead of Super Bowl ads — OpenAI's GPT-5.3-Codex sets new benchmarks in coding and general agentic capabilities, intensifying the AI enterprise market competition.
• New Codex features include the ability to use your computer in the background — OpenAI's updated Codex app introduces background task execution and an in-app browser, advancing its "super app" vision.
• Anthropic's Opus 4.6 and OpenAI's Codex: New AI Models Compete — AI competition intensifies with new agent team capabilities, expanded context windows, and enterprise platforms.
• [AINews] GPT 5.5 and OpenAI Codex Superapp — AI models are advancing in intelligence and cost-efficiency, driving a shift towards agentic workflows and open-source competition.
• OpenAI's Codex update lets agents build interactive enterprise workspaces via Sites and role-specific plugins — Codex's update transforms it into an agentic AI operating environment for business, enabling precise automation and interactive app creation.
• OpenAI Codex App: death of the VSCode fork, multitasking worktrees, Skills Automations — Agent-native coding interfaces and specialized open models are rapidly advancing, shifting developer workflows and evaluation paradigms.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by The Rundown AI.