Community

MJ

@mjackson

If you haven't tried Codex yet, you're missing something BIG.

Victor Mota

@vimota

OpenAI got me with the 2x increased usage limits on the Codex desktop app. It's a really nice experience. Super smooth - claude code app is so laggy. And the handoff experience is amazing. > I started a session locally on codex app > had to leave so I handed it off to codex cloud (with 2 attempts) > got back to my computer, compared both attempts, picked the better one, synced it back to my computer > continued to iterate on it locally

Derya Unutmaz, MD

@DeryaTR_

My new Sunday morning routine: 1. Get coffee 2. Check GPT-5.4 projects on the Codex App, continue & start new ones 4. Launch ChatGPT 5.4 Pro for fresh brainstorming sessions 5. Think/learn how to use the 90% of AI capabilities I have yet to explore 6. Drink more coffee

Evan You

@youyuxi

Did a full repo internal / public docs vs. source code alignment check + update - GPT 5.4 did a significantly better job than Opus 4.6. Note this is not a coding task but more like research / documentation. Same prompt but GPT 5.4's proposed revisions captured my intent much...

ashe

@ashebytes

given the interest, I productized my video hub! - have 1 place for video drafts + X, Youtube, tiktok vids - use the studio to comment + collab with others - get transcripts + add API keys for autosyncs/agent chat made with codex 5.3 🖤 linked below

John Jung

@johnjjung

Codex 5.3 xhigh spark has been phenomenal - you’re able to stay in a state of flow building. The focus and power of hitting problem after problem is incredible.

Mitchell Hashimoto

@mitchellh

Ahhhh, Codex 5.3 (xhigh) with a vague prompt just solved a bug that I and others have been struggling to fix for over 6 months. Other reasoning levels with Codex failed, Opus 4.6 failed. Cost $4.14 and 45 minutes.

Theo - t3.gg

@theo

Almost every verifiable problem I've been sent so far has been solved by 5.3 Codex. Please make sure 5.3 Codex can't solve your problem before you submit it to me. I'm a youtuber not a prompt box.

Mitchell Hashimoto

@mitchellh

I know this is pretty well established at this point, but Codex 5.3 is a much more effective model than Opus 4.6. I went back and forth on both for a bit, but haven't touched Opus at all now for a full week. First model to get me off of Opus... ever. Good job Codex team.

Mariusz Kurman

@mkurman88

Those who wrote "Try Codex" when I was hyping CC were right. Codex 5.3 is another level. It delivers so much with such high quality - I'm literally shocked. 5.2 was a mess; 5.3 is in another league.

Dennis Hannusch

@DennisHannusch

I started daily driving Codex with gpt-5.3-codex this week.. it's reaaally good. I've gotten used to complex workflows and context management, but Codex just does what I ask. I keep expecting quality to drop deep into a session, but it doesn't. @OpenAIDevs ya'll cooked!

ORO AI ✈️ ETHDenver

@getoro_xyz

OpenAI rolls out GPT-5.3-Codex with faster AI coding and measurable efficiency gains. Look at the chart. That Kurzweil Curve doesn't happen by accident. It's the result of a new data diet plan, shifting toward real-world and private workflow data.

Christopher Ehrlich

@ccccjjjjeeee

It actually worked! For the past couple of days I've been throwing 5.3-codex at the C codebase for SimCity (1989) to port it to TypeScript. Not reading any code, very little steering. Today I have SimCity running in the browser. I can't believe this new world we live in.

Dan Shipper 📧

@danshipper

BREAKING: At the Super Bowl halftime show, Bad Bunny says he prefers Opus 4.6 for vibe coding but turns to 5.3 Codex for gnarly engineering problems. "Opus pa' vibear, Codex pa' lo heavy" - he said

Guillermo Rauch

@rauchg

🆕 GPT 5.3 Codex (xhigh) achieves 90% on Next.js evals out of the box, "frame-mogging" the competition so to speak:

Angel ❄️

@Angaisb_

GPT-5.3 Codex is actually pretty insane with Three.js This Minecraft clone works smoothly and it didn't take too long to make I also tried Opus 4.6, but for some reason it got stuck

Angel ❄️

@Angaisb_

GPT-5.3-Codex with image gen skill (NBP) I'm loving this model so much. This is one-shot

Flavio Adamo

@flavioAd

Codex app is the most 'non-native' app that somehow feels more native than native apps

Pallav Agarwal

@pallavmac

Ok the new Codex app is incredible. You can build and run an Xcode project from Codex directly so you can vibecode iOS apps without opening Xcode!

Paul Solt

@PaulSolt

The Codex app feels like the missing IDE layer for the Codex CLI. One-click to open Xcode, run iOS/macOS apps, review diffs, and manage agents. Hard to go back after this.

Kristen Anderson

@FintechKristen

Because codex is the best for development. Everyone knows this. We aren’t using the other coding tools anymore.

Nate Berkopec

@nateberkopec

I have not seen a better code review agent that Codex 5.2 xhigh /review.

Anthony

@kr0der

Codex v0.91.0 has plan mode, and it's really thorough. this one plan took 27% of its context. if you've used Codex you know that 27% is a lot which means it's thoroughly searching before creating plans - no rushing/taking shortcuts. try it out by adding 'collaboration_modes =…

Rudrank Riyam

@rudrank

I spent 3 hours debugging and running all kinds of tests with Opus 4.5 for the discrepancy between CLI vs iOS app embeddings and the retrieval 5.2 Codex xHigh did "thinking" for like 20 minutes and casually one-shot it

Thomas Ricouard

@Dimillian

Yes. At this point, I don't care about speed anymore. Speed is solved by running multiple tasks in parallel. What I want is an agent that can think out of the box like a human, and Codex does just that.

Rudrank Riyam

@rudrank

Codex models is something you can just "trust" more than others I was in car and saw an issue created, and I let a cloud agent work on it for ~17 minutes

Henry

@henrytdowling

Codex is *really good* for coding, one day people will realize

Gavin Nelson

@Gavmn

Favorite recent use of Codex: - pull data from my Apple Music library with a two decade old mess of genre metadata - use MusicKit to pull in proper Apple Music genre metadata - write an Apple Script to update genres with the data from Apple's catalog 💆🏻‍♂️

Xeophon

@xeophon

Longest Codex run achieved with 8 hours 👀

Peter Steinberger 🦞

@steipete

codex is so good. point at messy PR, sizzles out the 3 fixes out of a big commit, cleanly separates it. done. also idk why it always says full gate but i fully adopted this.

Ian Nuttall

@iannuttall

One thing I have noticed with Codex CLI and 5.2 Codex high model recently I never worry about the context window or compaction any more It just does it, really fast, and the agent carries on like nothing happened, no loss in quality

Aaron Francis

@aarondfrancis

Codex just spent like 6 hours debugging CI while I played outside with my kids

AK

@ak_cozmo

codex hitting that inflection point where it's better at remembering API docs than i am. the shift from "AI suggests code" to "AI owns the implementation layer" is wild most underrated change: junior devs can now ship senior-level infra. distribution of execution ability is…

Numman Ali

@nummanali

Bloody hell, I'll say this GPT 5.2 Codex Extra High is a methodical beast It's updating the OpenCode OpenAI Codex OAuth plugin Literally not leaving any stone unturned This is the first model that feels like it's building for itself ie leaving the door open for future work