Codex @OpenAI

San Francisco
Joined November 2010
Turns out building with @ChatGPT Sites is pretty addicting. So is clicking cookies. 🍪 I made a little demo that does both, with a database and an MCP server so ChatGPT and Codex can join in. Come hang out. I hope it gives you an idea for something fun to build. Chapters: 00:00 A cookie clicker for humans and AI 00:27 The prompt I started with 00:41 Saving clicks in a database 01:16 Adding MCP tools 01:29 MCP in plain language 02:09 Testing with MCP Inspector 03:11 Asking ChatGPT for the click count 03:40 Clicking a cookie with Codex CLI 04:10 What will you build?
12
14
1
122
19,159
An update here: we’re upgrading all Codex for OSS grants to Pro 200! There will still be the 10,000 grants in this round, now at the higher tier. (If you've already received a grant, the upgrade should be reflected automatically.) Applications are still open :)
We started Codex for OSS six months ago. Today, we’re renewing the program with $100 Pro plans and doubling the number of grants (from 5,000 to 10,000) to reach more maintainers. (If you received a grant previously, we encourage you to re-apply!) openai.com/form/codex-for-os…
51
14
17
355
73,144
Crazy day to be the number 4.
1
6
1,254
pranav retweeted
Day 16/ 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 or resuming after a few hours and are now guaranteed to hit the cache.
You can now choose 12-hour prompt caching in the API. Persistent agents, like dots, reuse context throughout the day. You can keep shared instructions, tools, and context cached longer and pay the lower cached-input rate, even when requests are hours apart.
2,579
680
1,074
15,306
2,811,036
I'm excited to share that WebMCP is now enabled for MCP Apps in ChatGPT Desktop! 🚀 @OpenAI is also collaborating with modelcontextprotocol.io to make WebMCP the default way for agents to interact with MCP Apps. Another big step toward a more seamless agentic web!
36
25
11
336
30,732
LFG
Replying to @thsottiaux
🤫 we'll let the work do the talking
2
20
2,403
So far Dots > Muse Really like the Muse app but the model is dumb. Surfaces trivial emails and not smart enough to independently take on substantial projects Dots is smart enough for me to trust it
32
6
1
192
21,619
Building apps with GPT-Live-1? Give them a voice that fits. Here’s how to shape a voice with a prompt or choose from new international preset voices:
98
118
46
1,617
120,950
pranav retweeted
I built appshots for iOS, powered by dots + shortcuts. It has been life changing. I spend hours scrolling, reading and bookmarking things, then forget all of them. Now, I just say “Hey Siri, run dots” and keep going. Siri takes a screenshot and sends it to dots, which handles everything in the background. It finds the shoes the woman in the reel was wearing, signs up for a free gym class from an ad I saw, orders the book mentioned in the podcast, saves the video referenced in the x post to watch later, and more. dots reads the screenshot, and listens to my voice notes. Uses its own browser, to look up stuff: youtube videos, x posts on its computer. For walled apps like Instagram, I gave dots its own account, and it has been able to find the source post and from there find the products/places I am asking it to further research more accurately. Everything is saved to a ChatGPT Site and Google Doc, so I can look back anytime. An iOS shortcut can’t talk to dots directly. so it emails the screenshot to an inbox that dots checks regularly. The accuracy has been great. It finds exactly the right thing every time. It also is a much better solution than trying to use any bookmark tools within an app or a browser. Next, it would be so cool to have friends’ dots talking to each other, exchanging notes and building a shared feed.
30
14
5
239
37,200
icymi we've simplified API usage tiers! Build, Launch, and Grow, with each one unlocking higher TPM as your spend grows. And with the OAI Devs Plugin, building with the API is easier than ever - developers.openai.com/learn/…
4
802
we're dropping something heavy tomorrow @every it's the 8k word definitive history of Codex written from deep interviews with the team that did it including @gdb @thsottiaux @ajambrosino @embirico @mia_glaese @edbayes and more it'll also be agent-native so you and your agent can explore the full text of all the interviews in addition to reading the story :)
27
21
9
226
24,542
I took the new OpenAI Decisions API grocery shopping
5
1
26
1,736
pranav retweeted
At DevDay, we introduced ChatGPT Space: a new home for your team to collaborate with AI to get work done. In Space, you can create living documents with ChatGPT and the people you work with. Turn a messy conversation into an organized plan, and invite a teammate to edit together or leave feedback in comments. Tag ChatGPT or your dot to make changes or help with next steps. You can also ask ChatGPT to keep the page updated from your connected tools, so you can focus on the work—not updating the doc. See how it works below and go to chatgpt.com/space to get started. Available on Pro, Business, and Enterprise plans on web and desktop.
178
105
44
1,379
264,875
pranav retweeted
Day 9/ 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 heard enough feedback about power users running out of disk space.
4,500
439
639
15,011
2,748,403
Apple TV is the only part of Apple putting out hits currently
10
1,276
pranav retweeted
Replying to @rohit3a
Folks. Why do you think we picked 300k for context length for Codex. We ran the numbers and it’s better. We can run it at 1M too but it would just cost more of the usage. Good defaults are important.
331
63
85
3,342
521,593
Spotify is now completely useless as a tool for discovering new music and artists. When you start a radio based on a song you like, more than 50% of the radio is music you’ve already heard multiple times. Idk what they did, but it’s unusable now.
7
26
2,831
pranav retweeted
let's talk about submissions. since the beginning of the year, chatgpt plugin submissions have increased by almost 10x. then, since the explosion of interest in plugins after devday, they 3x'ed again! 🤯 it's been really difficult for our review team to keep up, and it shows in the review times. right now, the eta for you to hear back from us is about 2 weeks – that's too long! so, we've declared a temporary code red. we're relaxing some of our nice-to-have requirements so that we can increase throughput. it's going to take us about one week to catch up with our backlog, so if you have submitted so far you'll hear back in the next days. we're also: 1. going to show the eta in the submission portal so at least you know when you'll hear back from us 2. working on automating more review checks so you get immediate feedback without having to wait for our review team 3. taking another look at the requirements to make sure we check for only the most critical things to keep our users safe keep building plugins and keep submitting them; we're hard at work making that a great experiencer. thanks for bearing with the team. more context in the video:
70
10
13
277
41,022
🎯🎯🎯 Completely agree that we need more of this
If AI labs want personal assistants to take off at scale, they need to stop marketing vacations, reservations, party planning, etc. These are not problems. Focus on *the* problem facing Americans: the cost of living. Agents can offer a ton of help. Some ideas: (🧵)
2
45
5,686