OpenAI's superapp hiding inside Codex
OpenAI released a major Codex update transforming it from a coding agent into a comprehensive superapp with features like background computer use, parallel agents, and an in-app browser. The newsletter also covers Anthropic's new Claude Opus 4.7 model and OpenAI's first domain-specific model for life sciences.
Summary
OpenAI announced a significant update to its Codex platform, evolving it from a simple coding agent into a comprehensive superapp that integrates ChatGPT, Atlas, and Codex functionality. The new features include background computer use that allows Codex to operate Mac apps independently, parallel agent capabilities, an Atlas-powered in-app browser, memory retention across sessions, and inline image generation. Codex has reached 3 million weekly users with 70% month-over-month growth, with OpenAI explicitly stating they are 'building the super app out in the open.' This move positions OpenAI to compete directly with Anthropic's Claude Code and Cowork offerings. Separately, Anthropic released Claude Opus 4.7, which outperforms GPT-5.4 and Gemini 3.1 Pro on coding benchmarks, achieving 64.3% on SWE-bench Pro compared to its predecessor's 53.4%. However, it still trails behind Anthropic's unreleased Mythos Preview model at 77.8%. The newsletter also covers OpenAI's launch of GPT-Rosalind, their first domain-specific model for life sciences and drug discovery, which scored better than 95% of human scientists on RNA prediction tasks in blind tests. Additional updates include new AI tools like Windsurf 2.0 and HY-World 2.0, plus a practical guide for running AI models locally using Ollama.
About this episode
PLUS: Run an LLM on your laptop for free with Ollama
Key Insights
- OpenAI is explicitly building a superapp 'out in the open' with Codex as the foundation, transforming it from a coding agent into a comprehensive platform with computer use, parallel agents, and integrated browsing capabilities
- Anthropic is running two parallel development tracks with a fast 2-month public release cadence for models like Opus 4.7 and a gated frontier line in Mythos accessible only to exclusive partners
- OpenAI released two domain-specific models (GPT-5.4-Cyber and GPT-Rosalind) within three days, indicating a strategic shift toward purpose-built models for specific industries rather than relying solely on general-purpose flagship models
- GPT-Rosalind scored better than 95% of human scientists on RNA prediction tasks in blind testing by gene therapy lab Dyno Therapeutics, demonstrating significant capability in specialized scientific domains
- Codex achieved 3 million weekly users with 70% month-over-month growth, suggesting strong market adoption of AI-powered development tools
Topics
Transcript
Good morning, {{ first_name | AI enthusiasts }}. OpenAI has been teasing a superapp for months. Today, it shipped a major first piece. With a major Codex update bringing new features like background computer use, an in-app browser, parallel agents, and more, the rollout is OpenAI's clearest step yet toward the all-in-one platform it's been building out in the open. OpenAI’s superapp shift with Codex update Anthropic's Opus 4.7 tops rivals, trails Mythos Run an LLM on your laptop for free with Ollama OpenAI’s first science domain-specific model 4 new AI tools, community workflows, and more OPENAI The Rundown: OpenAI just updated its Codex platform, shifting it from a coding agent to a cohesive ChatGPT + Atlas + Codex app…
Full transcript available for MurmurCast members
Sign Up to AccessMore from The Rundown AI
Grok 4.6 storms the AI frontier
SpaceXAI's Grok 4.6 has emerged as a competitive frontier AI model, scoring 61 on benchmarks and undercutting rivals on price at 60% less cost, with Grok 4.7 teased for release in 3-4 weeks. The newsletter covers AI trends including practical guidance on using AI effectively, a personalized cancer vaccine startup for dogs built with AI tools, and updates on competing models and AI products.
Anthropic slips an invisible signature into Claude
Anthropic has announced the implementation of invisible watermarks in outputs from its Claude models to enhance transparency, aligning with the EU AI Act. The move has sparked mixed reactions, as stakeholders navigate the implications for open and private AI models.
Meta returns to its open-source roots
Meta is returning to its open-source roots with the launch of the Muse Glimmer model and plans for Muse Spark 1.2, positioning itself as a competitive force against China's AI dominance. OpenAI has also expanded its security program with a new hacking-tuned model, indicating a growing emphasis on cybersecurity in AI deployments.
OpenAI puts the safety brakes on Astra
OpenAI's Astra model, rumored to be GPT-6, is being treated as a potential 'critical' cyber risk following its success in solving significant problems, prompting increased safeguards and a delay in its rollout. This decision reflects broader concerns about AI's evolving capabilities and associated security risks as evidenced by recent breaches in the AI community.
AI designs viruses never seen in nature
Researchers used AI to design 16 novel viruses that infect only bacteria, successfully creating them in the lab to combat drug-resistant infections. While the work demonstrates beneficial applications, experts warn that the same AI tools could be repurposed to create dangerous pathogens, prompting calls for biosafety guardrails.