Tom

@simicvm @Dimillian This!

@Dimillian nice update! when will we get the ability to reference (@) files directly?

An update of ChatGPT for iOS is out with a bunch of new things and fixes for Remote! learn.chatgpt.com/docs/changelog…

@Dimillian @SIGKITTEN Remote has become GOAT, I use it all the time. Great work guys! The only thing I’m really still missing is push notifications, but you already know that 😅

@JustLingonberry @thsottiaux @thsottiaux sorry but this is real, please do something! 😩

@thsottiaux been loving the progress and the resets but its clear that usage is down across the board for everyone. I can barely get anything done like I used to without burning through usage. $200/month for a year and it has never been this bad. Please fix. Everybody is upset.

@simpsoka @SIGKITTEN @Dimillian Hahah well, since you’re on it, just an idea…

@simpsoka Kath, can you please give @SIGKITTEN or @Dimillian an ear pulling for not fixing mobile codex / remote push notifications? 😅

More papercuts fixes coming at ya!

@evi77ain @OpenAIDevs @thsottiaux @reach_vb It’s definitely not the fastest according to benchmarks. It’s the cheapest, but can take longer to achieve its goal.

Did I get this right? We have GPT-5.6 Luna. We can select it as the main agent. But spawn_agent literally can't use it for a subagent? The fastest model is unavailable for the exact role where speed matters most. This feels like a bad joke, @OpenAIDevs @thsottiaux @reach_vb 🥹

@simpsoka @romainhuet You delivered! It’s fixed now. Thnx!

@TomasPuelma @romainhuet Sure. Feel free to dm me what you’re experiencing too

Few things you might've missed in the announcement last week. We added an easier way to review PRs, edit code in Codex directly, and improvements to Computer Use. I love being able to quickly edit code in the diff -- I do this for adding small things, or debugging so I don't have to use my tokens for that.

@thsottiaux Banked reset please! The 10M milestone is supposed to be the last one to get resets, so is worth a banked one IMO.

10M! New day, new usage reset for paid users of Codex and ChatGPT Work. Lands in the next hour. Enjoy.

@da4004ad @thsottiaux I think the resets were until reaching 10M, so this was probably the last one 😢 hopefully not.

@thsottiaux Reaching the entire Swedish population size now. Has anyone extrapolated when the 11M reset will occur?

@Dimillian Thomas, sorry for the unrelated reply, but can you please tell codex to fix “remote” push notifications? They still almost never work for me. Also, a suggestion: a “Notify me” option for threads you want to receive PNs.

I need more stuff on my desk

@reach_vb @dkundel But no skills 🫠

With ChatGPT Work, you can kick off cloud tasks directly from the desktop or mobile app. It can use your configured plugins to work across your repos, docs, Slack, and the other tools you already use. I use it constantly from my phone: hand off a task, let it run asynchronously in the cloud, and come back when the work is done.

@gdb But no desktop skills 😢🫠

one of the best features of ChatGPT Work is that it runs in the cloud, meaning that it works from mobile, with your laptop closed. kinda crazy how long the main way to get the magic of agents has been while leaving your laptop cracked open!

OpenAI team, will we ever get the extended ~353k context window back in Codex? 🥺 @thsottiaux @ajambrosino @Dimillian @simpsoka @jxnlco @OpenAIDevs @OpenAI

@LexnLin @thsottiaux Yes!

I need a slow mode in Codex that costs 0.33x normal speed Would love to run tasks that take around 4h on normal speed overnight but with slow mode And a bit more clarity on how much fast uses compared to Standard would be nice :) @thsottiaux

@simpsoka @romainhuet It's really straightforward to reproduce, but hard to explain, just try to edit any txt file. You edit and it doesn't refresh onscreen, but it is editing on the background, and what's worse, it shuffles some keystrokes. Really messy.

@thsottiaux Give a banked reset, but with a two months expiration period instead of one

Embarrassment of riches. But looks like we might hit 9M soon. Should we reset the ChatGPT Work and Codex usage again or give it some space?

@LLMJunky @OpenAIDevs Nice! 👏

Hello fresh TL, moots, and builders. It's a pleasure to congregate here with all you fine legends this Tuesday morn. Welcome back! I've gotten the opportunity to spend some time with Sol, and it has been an absolute joy to use. But did you know that @OpenAIDevs suggest prompting 5.6 Sol differently than other models in the past? The evolution of coding with AI has truly entered a new era. We started with autocomplete, tabbing to finish a line or two of code. The models got better, and we started delegating entire tasks. The models got better once again, and we're on to building specs, and then subagents to implement them end-to-end. But now, we're in the era of outcomes. Rather than obsessing over every detail of what the model does along its path, your attention is best served on describing the end result: outcomes, expectations, evidence, guardrails, and what it really means to be "done." It doesn't mean that you're completely letting go of the wheel, but it can lead to better outcomes if you simply allow the model the autonomy to explore on its own. To help you navigate this shift, I created a meta prompting skill specifically for GPT-5.6 based directly on OpenAI's 5.6 Sol Prompting guidelines. Links in comments.

@Dimillian @PaulSolt @PhilippSpiess Awesome! 👏 Thomas, this “thinking” UX could be improved imo. Just papercut, but still.

If you update to the latest ChatGPT iOS app version, we have a very cool new feature from @PhilippSpiess. Codex visualisation now also works on iOS; you can ask for all kinds of graphs and custom content to be built on the fly!

@iruletheworldmo @thsottiaux @AniBaddepudi It’s 960 again

@iruletheworldmo @thsottiaux @AniBaddepudi Can confirm

Updates for Codex and ChatGPT Work users. No nerfing, only good stuff! - We have landed inference optimizations and are passing down savings to all the subscriptions for GPT-5.6 Sol. That should result in around 10% more usage on its own. - We noticed that by changing the context size limit in the product to 372k for GPT-5.6 Sol, up from 272k for GPT-5.5, it resulted in more usage being charged than intended. We have reverted to 272k and will work to roll back out to 372k in the days to come. You should notice that usage drains significantly less after this change. - To understand where the extra usage was coming from, we ran some experiments where reasoning efforts were changed (referred to as juice values under the hood) and have reverted this. - There is slightly more usage of multi-agent than intended in high and xhigh reasoning effort, we are fixing this going forward. Also fixing a small other thing we noticed with auto-review where we can be more efficient. And we continue to have the 5h limit temporarily not apply. Enjoy the rest of the weekend!