Hey there!
Summary of this email:
- 🗞️ News of the week: what matters to us from OpenAI's DevDay 2026
- Improve your agents' Harness
- Café con Codely recap (watch it here)
- The joke of the week
🗞️ News of the week: OpenAI launches its dots agents and GPT-6.1 Sol at half the price of Opus 5.5
On Tuesday OpenAI held its DevDay 2026 with more than 20 announcements across ChatGPT, Codex and the API.
Out of everything they presented, we highlight 3:
🤖 Dots, your personal agent
One of the most striking announcements is dots: agents that can keep working on your goals even when you are not there.
Each dot has its own computer and browser in the cloud and can run work in parallel with other agents. You can interact with it from ChatGPT and also integrate it into Slack or Teams.
The example for devs is especially interesting: it can review customer feedback, spot recurring bugs and end up opening a finished PR with a video that shows the change. Something Cursor added a few months ago and we loved.
You can also call it to interact by voice. Both from your phone and from your computer.
We think it is a very good answer to Meta Muse and Grok Bot, and if you already use ChatGPT/Codex, it makes total sense to try it.
We have had access to it and we are using it in depth. Tomorrow we will be live at 9 CEST on Café con Codely sharing our opinions. On our YouTube, Twitch and X.
💸 GPT-6.1 Sol: almost Astra for a fifth of the price
According to OpenAI, GPT-6.1 Sol gets close to Astra in agentic coding and computer use for a fifth of the price per token.
In the API it costs 2 dollars per million input tokens, 10 for output and 0.10 for cache reads. That is exactly half of Opus 5.5, which costs 4, 20 and 0.20.
It is already available in Codex and in the API as gpt-6.1-sol.
🎯 Decisions API: OpenAI wants its own Jev too
In the AI era we are living in, you cannot save anything for an announcement day.
OpenAI probably had this Decisions API ready before Jev launched, but they kept it for their DevDay.
Since Jev came out first, it got the novelty factor and everyone started talking about it.
Now that OpenAI's Decisions API is out, it is one of the least discussed things from DevDay.
The Decisions API uses Luna to answer questions with a closed set of answers that you define. You give it text or, unlike Jev, images, and it returns an option to classify, route or pick an agent's next step.
It is in limited preview and they will open it to everyone in the coming days.
🤖 Improve your agents' Harness
We recently published a free email course that is 5 emails long.
In the course we explain how, with specific skills, you can greatly improve the Agent Harness of your projects.
You can sign up from its own landing page.
We hope you like it!
☕ Café con Codely recap
This is the top 3 of headlines we discussed in the latest Café:
- Opus 5.5: Fable 5.1 level, 30% faster and 40% cheaper than Opus 5. Claude Code increases the 5-hour window limit by 20% and gives you a limit reset to try the model until October 22.
- The senior engineer death spiral: fewer megaprojects, more quick wins. Sunil Pai recommends helping the team with small tasks and sharing your work instead of isolating yourself in a huge project.
- Jev in Postgres: natural language search without indexes or embeddings. The pg-jev extension adds a
jev()function to theWHEREclause and sends the rows to Jev in batches of 20 after it applies the deterministic filters.
You have the full Café recap with links to each story at the moment we discussed it.
Tomorrow we will be live at 9 CEST discussing this week's news on Café con Codely. On our YouTube, Twitch and X. 🙌
And since you've made it to this part of the newsletter, here's the joke of the week, which I know you were waiting for:
Why is Redis the most respected database? Because it has a lot of cache! 😂 😂 😂
Cheers!