GPT-5.1 represents a meaningful step forward in LLM capabilities. Three key improvements stand out:
1. Engine Segmentation & Personality Presets
The ability to segment different engine types with distinct personalities is genuinely useful. As a GTM builder, this means I can deploy contextually-optimized responses without extensive prompt engineering overhead.
2. Superior Instruction Following
The model now handles multi-step constraints simultaneously. Complex instructions that previously required 3-4 iterations now work on the first try. This directly reduces latency in production systems.
3. Improved Tone Adaptation
GPT-5.1 understands conversational context better. It shifts tone appropriately based on input, which matters more than people realize for enterprise adoption. Technical superiority loses to human-like interaction every time.
The Real Unlock: This isn't a revolutionary leap. It's a solid incremental advance that compounds when deployed at scale. The real advantage goes to teams building on top of this—not those claiming AGI is here.
The moment agentic coding tools got a proper async loop was the moment mobile access became non-negotiable.
What it is: Codex can now be accessed and managed from the ChatGPT mobile app on iOS and Android, connecting your phone to any environment where Codex is running.
Most agentic tools assume you're sitting in front of your computer. Codex often isn't fast work. It writes features, runs tests, proposes pull requests, and sometimes hits a wall waiting for your input. The useful part of this update is that your phone becomes a legitimate control surface, not a stripped-down viewer.
What makes it different: When you connect to any of your machines where Codex is running, the app loads the live state from that environment so you can work fluidly across active threads, approvals, plugins, and project context. The session state stays synced via a relay layer that keeps machines reachable without exposing them to the public internet.
Key features:
QR code pairing between the ChatGPT mobile app and Codex on Mac (Windows support coming)
Full thread management: start, steer, unblock, approve from phone
Computer Use lets Codex operate desktop apps and logged-in websites through Chrome, all accessible from mobile
Available on all plans including Free and Go, in preview
Enterprise additions: programmatic access tokens and HIPAA compliance ship alongside
Benefits:
Close the feedback loop on long-running agent tasks without being at your desk
Approve or redirect Codex before it compounds on a wrong assumption
No new app to install: surfaces inside the ChatGPT app you already use
Who it's for: Engineers and technical teams using Codex for production-grade coding tasks who need to stay close to ongoing agent work across devices.
My read is that mobile parity for agentic dev tools is becoming table stakes, and this is a solid implementation of it.
Monitoring long-running agent tasks from a phone is a huge quality-of-life upgrade. Can you actually trigger a 'stop' or 'retry' command directly from the mobile interface if a task goes off the rails?