OpenAI DevDay 2026

What OpenAI announced at DevDay 2026, and what people made of it.

On 29 September OpenAI announced more than twenty things at DevDay in San Francisco: dots, always-on agents with a computer of their own; GPT-6.1 Sol, which OpenAI says comes close to Astra at a fifth of the price; ChatGPT Space and Pages for teams working alongside agents; Codex in the cloud, computer use in the Agents API, and a Pro 500 plan. We are collecting the launches in OpenAI's own words, what was shown on stage, what people tried and argued about in the first hours, and the night before, each in the poster's own words. Nobody in these bylines is on eChai, and every entry links back to where it was posted.

Founder Notes →

Curated from OpenAI 29 September 2026
DevDay 2026 Recap →

Image from OpenAI

OpenAI's own list of what it launched at DevDay on 29 September, in five parts. New ways of working: dots, always-on agents powered by GPT-6 Astra with a cloud computer of their own, on Pro and Business Premium first; GPT-6.1 Sol, which OpenAI says comes close to Astra at a fifth of its price; Ultrafast, a speed tier up to eight times faster in Codex; and Private Intelligence for businesses. For developers: Codex in the cloud, a Codex CLI you can steer by voice, code review in the desktop app, Codex Security Cloud, a Decisions API built on Luna, and computer use in the Agents API, which Amazon also runs as Bedrock Managed Agents. For plugins: extensions that live in ChatGPT's sidebar, a new submission flow, plugins inside Sites, and MCP events that start automations. For teams: ChatGPT Space, Pages, shared slides, team tasks, @ChatGPT in Slack and Teams, a Meetings plugin and shareable profiles. For subscribers: Sign in with ChatGPT across 16 partner tools, a Pro 500 tier at 25 times the Plus allowance, and an OpenAI Marketplace where an enterprise's OpenAI commitment can pay for partner software.

From OpenAI

DevDay 2026 is our biggest yet, with more than 20 major announcements across ChatGPT, Codex, our models, and entirely new forms of working with AI.

Today, we introduced agents that can take on ongoing responsibilities and new ways for people and AI to work together. We also expanded our commitment to an open ecosystem by opening up ChatGPT as a shared surface where humans and agents can collaborate and where developers can directly launch new native experiences to our collective 1.2B weekly users.

Read it in full on OpenAI →
Curated from X →

klöss on where the anger over the plans actually came from, at the end of the day.

klöss 3w

people aren't mad OpenAI launched a $500 plan

they're mad Tibo spent weeks hyping GPT-6.1 Astra like it would solve all your life's problems, just to then not ship it

they're mad GPT-6.1 Sol dropped one week after GPT-6 Sol (rushed???), and most devs are still reaching for Opus 5.5

they're mad the team knew today wouldn't land without GPT-6.1 Astra yet hyped it like the second coming of Jesus

and they're mad OpenAI sold us a 20x plan for $200, cut it to 10x, then asked for $500 to get 25x on the same day Tibo said "a lot of compute is online"

want a crash course in max aura loss?

study OpenAI today
Curated from X →

Alexander Embiricos, who works on Codex at OpenAI, on how far the rollout had reached by the end of the day.

Alexander Embiricos 3w

Dots rollout update:
- We are out to all of Pro500, Pro200
- Just starting Pro100
- We're a little delayed for Business Premium and Enterprise beta. Looking like tomorrow, sorry!
Curated from X →

cova, who works on art and technology at OpenAI, with a game made on the handheld the same evening.

cova 3w

i made a game on the devday gameboy!

inspired by @casey's nyc rides 🍕
Curated from X →

Gavin Nelson, a designer at OpenAI, on a small feature in the dots app.

Gavin Nelson 3w

one of my favorite little features we shipped in dots: with the app open, just raise your phone to your ear to call your dot

(you'll need to toggle this on in settings, voice)
Curated from X →

Tenobrus after a first evening with dots: the fullest list of what did not work yet, in a post that also calls it the agentic form factor people have been waiting for.

Tenobrus 3w

overall this feels somewhat underbaked, like they had to release this early to compete with muse. just a lot of little points of friction and weird bugs in onboarding and in the general experience

- you can only start a new dot on a desktop browser, not on your phone?? insane choice
- when i did go to start one the dot color / profile picker was broken, like full on would not let me edit anything until i cleared browser cache
- frequently quite slow with little feedback or indication about what the model is doing
- some strange overrefusals and double and triple checking for approvals on things im directly asking it to do
- the cloud browser gets blocked signing into most sites bc of cloudflare anti-bot checks, which muse is presumably pulling some shit to get around
- no ability to... open multiple chats with the same dot or have multiple dots? both of which seem quite confusing.
- the calling experience is honestly p bad, I find it very off-putting that they decided to emulate a phone call and take over the entire screen in the phone app, and openai voice models try way too hard to sound "natural" in a way that often is just glitchy
- it won't really tell you any details about how its memory or configuration works, there's no agents.md you can edit if you know to ask, you just kinda have to... trust their proprietary memory system is learning the right things
- there's this whole new "space" concept meant for multiplayer collaboration, but as far as i can tell no native way to just have ur dot message someone else's dot or set up a group chat. like u literally would have to use slack for that to work

i say all these things w love, bc i do think this is an incredibly powerful base . im sure openai will continue to build and improve this, quite quickly, and i genuinely think this is indeed the agentic form factor we've all been looking for a while now. but it is *not* quite yet "magical"
Show more
Curated from X →

TechHorn Lab on Manus, which the post says launched Manus 2.0 and a personal assistant called Cue the day before DevDay.

TechHorn Lab 3w

You have to feel for Manus.

It launched Manus 2.0 and its personal assistant Cue on September 28. Just one day later, OpenAI showed up with ChatGPT Space and dots.

One lets people and AI work together in a shared workspace. The other has its own cloud computer and can work for you around the clock. Seeing them together, it’s hard not to think: aren’t these basically OpenAI’s versions of Manus and Cue?

Manus helped kick off the Agent frenzy when it first appeared. Now, even a major update struggles to make the same splash when the giants are competing for the same users.

Perhaps Manus Studio still has something fresh to offer: while AI makes a video, you can jump into the timeline, insert your own footage, swap the music and adjust animations whenever you want.

#Manus #OpenAI #AIAgents
Curated from X →

Tibo on what the first days of dots are for, before teams of them are offered.

Tibo 3w

We will have a few million dots online within days, working on all sorts of things across such a diverse and large community.

Excited to learn from all of you on what you love and what doesn’t yet feel magical. Personally I felt a jump after 2-3 days of use after teaching it more about my preferences and things on my mind. It learns very quickly to be most useful and it can take on surprisingly ambitious tasks on its own.

We’re learning from how you all use your primary dot before releasing the ability to create an entire team of them.
Curated from X →

Tibo answering a Community Note on the dots launch. The two sources do not say the same thing yet: OpenAI's help page says dots will not count against a plan's allowance for the next month and that usage terms come after that, and this post says a primary dot's own work will always be included, with only the Codex tasks it starts drawing on usage and more speed a paid extra later.

Tibo 3w

I got community noted, but the note is wrong 😅

Your primary dot is included in your plan and will be available 24/7. If you ask it to create a codex task for you, that one will be drawing usage as usual. But when your dot does work directly, that uses nothing and is all on top of your plans usage. In the future you will be able to increase the speed and allow your dot to have more bandwidth, this will be a paid feature, but the baseline functionality will always just be included in your plan.
Curated from X →

OpenAI's chief financial officer, Sarah Friar, on CNBC the afternoon of DevDay. The clip is nine seconds.

Wall St Engine 3w

OpenAI CFO Sarah Friar accidentally calls Dots “Muse” on CNBC:
Curated from X →

Dmitriy Kovalenko, who works at OpenAI, answering a thread that said Codex writes diagnostic logs to disk without stopping.

Dmitriy Kovalenko 3w

I also have a little announcement for DevDay. I asked Tibo to include it in the presentation but he didn't read my slack DM 😭

My first onboarding project is merged. I fixed this issue once and forever! I reduced amount of disk writes by 50% and the size of stored files by 4x
Curated from X →

Siqi Chen, who builds cfo.ai, on 6.1 Sol in the company's own evaluations.

Siqi Chen 3w

we found opus 5.5 to be 2X as efficient as gpt-6-sol in our internal https://cfo.ai evals so i was not expecting 6.1-sol to improve on that much

but to see it 3X the efficiency of opus 5.5 and 2.5X against even sonnet 5.5 is actually 🤯

this is an insane point release
Curated from X →

Pauline P. Narvas, in developer experience at OpenAI, who moderated the closing Q&A.

Pauline P. Narvas 3w

I’ve done a lot of cool things in my career, but moderating the closing DevDay Q&A with these legends is definitely up there.

Thank you, @sama, @thsottiaux, @tejalpatwardhan, @CoryDecareaux, @romainhuet and our community for the trust 💙
Curated from X →

Tibo on TBPN explaining how computer use got ten times faster in a year: the model and the harness together, and a second agent that watches the first.

TBPN 3w

OpenAI made computer use 10x faster in a year.

@thsottiaux breaks down what changed under the hood:

“It’s not that we made the model sample faster. It’s both the harness and the model.”

“How much does it need to think before it can take the next action reliably? That’s one element that we really improved.”

“We have a second agent, which we call AutoReview, or internally we call it Guardian, which watches over the primary agent.”

“So we also made this AutoReview model faster.”

“We were able to use Astra at scale in order to improve the performance of the harness, do more things in parallel, remove little stalls.”

“Combine all of this together, and then we got this 10x speedup.”
Curated from X →

Maddie D. Reese on a detail from the closing Q&A, with the list of tools a dot said its own computer carries.

Maddie D. Reese 3w

Fun note from Tibo at the closing q&a of DevDay:

dots get their own Linux cloud computer, and that cloud computer comes preloaded with apps like Blender!

I checked with my dot to get a fuller list of the preloaded apps and tools

More entries

eChai Partner Brands The partnership →