
OpenAI DevDay 2026 Recap for Developers
Quick Answer
At OpenAI DevDay 2026, the company unveiled GPT-6.1 Sol, a cost-effective model with Astra-level performance, alongside new APIs for agents and cloud-based Codex environments, enhancing developer capabilities for persistent workflows.
Quick Take
The Decisions API and expanded plugin system further support automation and collaboration across applications.
Key Points
- GPT-6.1 Sol offers Astra-level performance at one-fifth the cost of Astra's token pricing.
- The Agents API now supports computer use, enabling software operation through graphical interfaces.
- Codex can run in cloud environments, allowing remote task initiation from various devices.
- The Decisions API uses the Luna model for classification and request routing tasks.
- ChatGPT's plugin system now supports sidebar experiences and event-driven automations.
📖 Reader Mode
~3 min readOpenAI announced a series of product and developer updates at DevDay 2026, including GPT-6.1 Sol, computer use for the Agents API, cloud-based Codex environments, a Decisions API, and new plugin capabilities for ChatGPT.
For developers building agents, the Agents API now supports computer use, allowing applications to operate software through graphical interfaces. The API also incorporates multi-agent capabilities from Codex, tool search, tool calling, and context compaction, with OpenAI managing the underlying execution infrastructure. The functionality is available through the API and in Codex and ChatGPT Work for selected plans.
OpenAI also released GPT-6.1 Sol, an update to GPT-6 Sol aimed at coding, computer use, and professional tasks. OpenAI says the model approaches GPT-6 Astra on several evaluations while charging one-fifth of Astra's standard input and output token prices. Cached input costs $0.10 per million tokens. GPT-6.1 Sol is available through the API, ChatGPT Work, and Codex.
Codex can now run in cloud environments in addition to local computers, allowing developers to start remote tasks from other devices. OpenAI also updated the Codex CLI with voice input and an /agents interface for delegating and monitoring multiple tasks. A new code-review workflow can analyze diffs and potential issues in GitHub pull requests and GitLab merge requests, while Codex Security Cloud can scan repositories and new commits, investigate findings, remove duplicates, and prepare fixes.
Another developer release is the Decisions API, currently in limited preview. It uses the smaller Luna model to select from a predefined set of answers based on text or image context. OpenAI positions the API for tasks such as classification, request routing, and choosing an agent's next action.
OpenAI also expanded ChatGPT's plugin system. Developers can now build sidebar experiences, interactive conversation panels, and custom file viewers. Plugins can also respond to events through the proposed MCP Events specification, allowing automations to start when an event occurs in a connected application.
DevDay also introduced Dots, persistent agents that can work on ongoing tasks, and ChatGPT Space, a shared workspace where teams and agents can work with common context. Together with the developer releases, these changes extend OpenAI's platform from individual model calls toward persistent agents that can use tools, operate software, collaborate, and execute tasks remotely.
Community discussion focused on the growing emphasis on agents and developer infrastructure. Developers highlighted computer use, cloud-based Codex, and lower-cost models as practical additions, while others questioned whether the releases represented significant advances over capabilities already available across existing agent platforms.
OpenAI developer Victor Nunez shared:
You would not believe just how much better dots got in a single week pre DevDay. I expect this to become the favorite way of interacting with AI for most people.
Meanwhile developer @TokenGremlin posted:
Thinking about it more, I actually think GPT-6.1 Sol might be THE BEST release in OpenAI’s history. It gave me a really unusual feeling that I almost never get from a model: Luna pricing, Astra-level quality.
DevDay 2026 showed OpenAI moving beyond individual model interactions toward persistent agent workflows. With autonomous agents, lower-cost models, cloud-based coding, and expanded ChatGPT integrations, the company is building infrastructure for agents that can work continuously across applications and tasks.
About the Author
Daniel Dominguez
Show moreShow less
— Originally published at infoq.com
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from InfoQ AI, ML & Data Engineering
See more →Google Cloud Workbench Notebooks Extension Connects VS Code to Google Cloud's Jupyter Notebooks
The Google Cloud Workbench Notebooks extension for VS Code allows developers to seamlessly connect their local IDE to managed Jupyter notebook environments on Google Cloud, enhancing ML workflow efficiency. This integration eliminates context switching, enabling smooth transitions from local experimentation to high-performance cloud computing.

