Discover Codex’s Hidden Features: Browser Integration, High‑Power Model Mode, and More
The article walks through ten lesser‑known Codex settings—including browser capability, unlocking the highest inference strength, importing work from other AI apps, advanced computer control, quick‑snapshot shortcuts, the dormant GPT‑5.3‑Codex‑Spark mode, phone‑to‑Mac remote control, memory/Chronicle preview, and plugin tips—providing step‑by‑step paths and screenshots for each.
1. Browser Capability
Codex’s built‑in browser can import Chrome cookies and passwords, allowing the ChatGPT browser to act like native Chrome. All logged‑in sites become accessible without re‑entering credentials. Path: Settings → Browser → General → Import . Screenshots show the import screen and a sample site ("X") displaying logged‑in content.
2. Model Inference Strength
The front‑end default offers five inference levels; a hidden Maximum level can be enabled manually. Path: Settings → Configuration → Model → Available Inference Strength . A screenshot illustrates the toggle.
3. Import Work from Other AI Apps
This long‑standing feature now supports more content types. Path: Settings → General → Import Work from Other AI Apps . The author demonstrates importing Claude Code settings, plugins, Skills, projects, and chat history into Codex without friction, with accompanying screenshots.
4. Computer Control
Codex’s computer‑control abilities have been strengthened. The author recommends enabling all options, noting a recent test where an app was controlled with deep interaction, feeling like a GPT‑5.6‑powered robot typing and moving the mouse. Screenshots document the control panel and the test run.
5. Application Snapshot
A shortcut key can instantly send the frontmost window’s content to Codex, providing the simplest screenshot‑to‑dialogue workflow. The author shows the shortcut configuration and the resulting snapshot view.
6. Overlooked GPT‑5.3‑Codex‑Spark
When the quota is near exhaustion, the GPT‑5.3‑Codex‑Spark model remains viable. The author shares a tip: assign a complex task just before the quota runs out, keep the model set to GPT‑5.6 high, and increase speed to 1.5. Screenshots illustrate the setting.
7. Phone Remote Control & Inter‑Mac Linking
The author’s iPhone can control two Mac computers, and the Macs can control each other. Screenshots display the remote‑control interface and the inter‑Mac connection.
8. Memory & Chronicle Preview
Memory acts as a mutable work notebook to reduce repeated context provision. Path: Settings → Personalization → Memory . The author compares Memory with AGENTS.md in a table, highlighting differences in content, reliability, scope, edit method, and suitable use cases. Chronicle, a more aggressive feature, can generate local memory from recent screen content, reducing context prompts but consuming quota and introducing privacy and prompt‑injection risks.
9. Plugins
The author links to a previous article recommending many Codex plugins and notes that new plugins are frequently added, offering deep utility.
10. Call for Feedback
Readers who find the guide useful are encouraged to like, share, and comment.
Signed-in readers can open the original source through BestHub's protected redirect.
This article has been distilled and summarized from source material, then republished for learning and reference. If you believe it infringes your rights, please contactand we will review it promptly.
Old Zhang's AI Learning
AI practitioner specializing in large-model evaluation and on-premise deployment, agents, AI programming, Vibe Coding, general AI, and broader tech trends, with daily original technical articles.
How this landed with the community
Was this worth your time?
0 Comments
Thoughtful readers leave field notes, pushback, and hard-won operational detail here.
