Codex 28-day challenge: the day-by-day log
Tibo pledged publicly that for the next 28 days, every day would bring either a new improvement or a full reset of Codex. The pledge promises one or the other, not a reset every day. This page logs how each day was kept and uses it to estimate the odds of the next reset.
Codex 28-day challenge
Progress · Ended
- Days logged
- 28 / 28
- Improvement days
- 14
- Reset days
- 14
- Challenge ends
The 28-day challenge has ended. The full day-by-day log is kept below.
The pledge in his words
Over the next 28 days, each day we’ll either ship one thing that is a clear improvement and relevant for most codex/work users or ship a full reset. Let the improvements begin.
Day-by-day log
| Day | What happened | Details | Post |
|---|---|---|---|
| Day 1 | Improvement shipped | We have optimized the default speed to be ~50% faster across GPT-6 Astra and GPT-6.1 Sol through the subscription across all our products and partners using Sign in With ChatGPT… | View |
| Day 2 | Full reset | 1/ We have made Auto-review free for all users signed in through a ChatGPT account. You can enable it in settings > permissions > auto-review. Auto-review improves upon the… | View |
| Day 3 | Full reset | The big one is GPT-6 in Chat, but today is also a little celebration day with a new high of 40M active users across Codex and ChatGPT Work. Loading a banked reset in everyone's… | View |
| Day 4 | Improvement shipped | We have improved steering to be instant, leading to the model now reacting much faster to adjustments you make, allowing you to course-correct direction in realtime and not have… | View |
| Day 5 | Improvement shipped | Composer predictions in the desktop app. Often leads to a double take with how on point they are. Included in the Pro plans without consuming usage. | View |
| Day 6 | Full reset | Reset | View |
| Day 7 | Improvement shipped | We have increased default context length for GPT-6.1 Sol to 500k when used through Codex & Work, up from 272k or an ~ 80% increase. We will go higher in the future, but it helps… | View |
| Day 8 | Full reset | Reset | View |
| Day 9 | Improvement shipped | We've rolled out optimizations to how we store state on disk for the chatgpt & codex desktop app that should reduce the amount stored on disk by 60%. Small ship but we've… | View |
| Day 10 | Full reset | The reports of my reset button being lost are greatly exaggerated Day 10/ Reset | View |
| Day 11 | Improvement shipped | We've landed a series of changes and more compute to very significantly decrease the occurrence of the disruptive "Selected model is at capacity. Please try a different model"… | View |
| Day 12 | Improvement shipped | We’ve made improvements to Codex & Work that increase the effective usage you get. We achieved this by reusing context across subagents, keeping context cached longer, and… | View |
| Day 13 | Full reset | Le limit reset https://t.co/zXd2DIH5xf | View |
| Day 14 | Improvement shipped | Small quality of life improvement. You can now edit previous messages within a thread in the desktop app. Long time requested, and took way too long to get that one out. | View |
| Day 15 | Full reset | Nevermind, we're still cooking... and reset. Enjoy. | View |
| Day 16 | Improvement shipped | You can now cache for 12h in the API and for logged in users across Codex & Work we keep cache forever and you don’t have any penalty for coming back to an old conversation… | View |
| Day 17 | Full reset | Bet you thought we forgot, but no, it’s reset day today. Also good time to remind you to get Daybreak Blue access, it only takes 15 mins and you don’t want to have to do it in… | View |
| Day 18 | Full reset | Decisions API is GA, OpenResponses joined the Linux Foundation, we shipped some performance improvements and Uber is going all-in on codex & work. But yes… that’s a reset! | View |
| Day 19 | Improvement shipped | Small updates. Calls connect much faster for dots, changing reasoning in codex & work mid conversation does not invalidate cache anymore and a we've also updated the codex cli. | View |
| Day 20 | Full reset | Rest and reset | View |
| Day 21 | Full reset | Wonder what happens the remaining seven days. But today is a reset. | View |
| Day 22 | Improvement shipped | Terra is back! Our first model that is significantly better at writing, personality and nuanced understanding of requests. Packs a punch across the board, very good at computer… | View |
| Day 23 | Full reset | Much requested iCloud Main & Calendar plugins. Just kidding, today is a reset. Still much requested though! | View |
| Day 24 | Improvement shipped | Text your dot just like anyone else. Makes it super easy to share/forward things, leave it voice notes or you can even go creative with "Hey Siri, text Iggy to <INSERT WILD… | View |
| Day 25 | Improvement shipped | We are doubling all usage limits for all Pro 100/200/500 tiers and Business Premium, effective today. We heard your kind feedback and won’t be reducing usage limits for Pro 200… | View |
| Day 26 | Improvement shipped | You can now easily switch between multiple accounts on mobile, and coming soon to desktop too. Next two days will be resets… banked… do with this what you may 👀… | View |
| Day 27 | Full reset | Banked 1/2 | View |
| Day 28 | Full reset | Banked 2/2 | View |
How it changes the odds of a reset
During the challenge the estimate is made of two parts added together. One is the usual chance of a reset, at the reset rate of the 15 days before the challenge began. The other is the extra chance the challenge adds each day. Whether the next day is a reset day depends on the day before: Tibo rarely does two direct resets on consecutive days, and a second reset day in a row usually comes as a banked reset. So the chance after a reset day and the chance after an improvement day are counted separately, each learned from the days logged so far. Because the two parts are added together, the odds during the challenge are higher than usual as long as no hint from Tibo is in play. Both rates, the usual one and the one during the challenge, are shown below so you can compare them.
Current estimate: 12% in the next 24 hours, 33% within 3 days.
Usually about one reset every 7.5 days; during the challenge, about one every 7.5 days.
based on 2 resets in the last 15 days
FAQ
Does the 28-day challenge mean a reset every day?
No. In Tibo's words, every day brings either an improvement or a full reset. A reset is just one way to keep the pledge.
When does the 28-day challenge start?
It starts with the post where Tibo made the pledge. The end date shown on this page is counted as 28 days from there.
How is the reset probability calculated?
We take the usual reset rate and add the extra chance the challenge brings each day; how large that chance is depends on whether the day before was a reset day or an improvement day, and on how many reset days followed each kind of day among the logged days. Tibo's latest posts then add a limited adjustment, which can push the odds up or down. The method is on the methodology page.
How often is the data updated?
The collector checks Tibo's latest posts about once a minute, and a newly logged day shows up here within a few minutes.