OpenAI takes aim at Anthropic with beefed-up Codex that gives it more power over your desktop

2026-04-16 · Source: AI News & Artificial Intelligence | TechCrunch · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering, Robotics & Autonomous Systems · Depth: Novice, short

Summary

OpenAI has announced a significant revamp of its automated coding tool, Codex, introducing several new features to enhance its capabilities and competitiveness against rivals like Anthropic's Claude Code. The most notable update is an "agentic" function that allows Codex to operate in the background on a user's Mac, opening applications, clicking, and typing, effectively deploying multiple agents in parallel without interfering with other user tasks. This enables Codex to assist with auxiliary coding tasks such as iterating on frontend changes or testing apps. Additionally, Codex now includes an in-app browser for web application commands, a "memory" feature to recall past work sessions, and new image-generation abilities for creating product concepts and mockups. OpenAI has also integrated 111 plug-ins from apps like CodeRabbit and GitLab Issues, expanding Codex's ability to handle clerical work. A new pay-as-you-go pricing option is available for ChatGPT enterprise and business customers.

Key takeaway

For AI Engineers evaluating coding assistants, OpenAI's revamped Codex, with its new agentic background operation and 111 plug-in integrations, presents a compelling alternative to Anthropic's Claude Code. You should explore its capabilities for automating auxiliary coding tasks, managing web applications, and generating visual assets, especially if your team uses ChatGPT enterprise or business and can benefit from the new pay-as-you-go pricing.

Key insights

OpenAI's revamped Codex introduces agentic background operation and expanded integrations to compete with Anthropic's Claude Code.

Principles

• Agentic AI can operate in parallel with human users.
• Browser integration expands AI tool control over web apps.

Method

Codex agents operate by controlling desktop applications via cursor actions, running in the background to perform auxiliary tasks without user interruption.

In practice

• Automate frontend changes and app testing with background agents.
• Generate product concepts and mockups using new image features.

Topics

• OpenAI Codex
• Anthropic Claude Code
• AI Coding Tools
• Agentic AI
• Desktop Automation

Best for: CTO, AI Engineer, Machine Learning Engineer, Tech Journalist, Director of AI/ML, AI Product Manager

Related on AIssential

• OpenAI Boosts Codex's Core Functionality — OpenAI's Codex updates enhance desktop control and multi-agent capabilities, directly challenging Anthropic's offerings.
• Anthropic just shipped an OpenClaw killer called Claude Code Channels, letting you message it over Telegram and Discord — Anthropic's Claude Code Channels offers asynchronous AI agent interaction via messaging apps, challenging open-source alternatives.
• 😸 Anthropic and OpenAI Both Dropped Their Best AI Models On The Same Day — Leading AI developers are advancing models towards autonomous job execution, emphasizing either broad utility or deep specialization.
• OpenAI takes Codex mobile — OpenAI's Codex mobile integration enhances developer workflow by enabling remote management of AI coding agents.
• How to Maximize OpenAI’s Codex — OpenAI's Codex, particularly with GPT-5.5, offers speed and precision for coding tasks, rivaling Claude Code.
• Anthropic Introduces Agent-Based Code Review for Claude Code — Anthropic's agent-based code review system uses multiple AI agents for deeper, parallel analysis of pull requests.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AI News & Artificial Intelligence | TechCrunch.