OpenAI DevDay 2026 Recap for Developers
C · Weak signal InfoQ news recap that relays OpenAI's DevDay 2026 announcements with a few community reactions; it adds no facts beyond the vendor's own releases.
Summary based on the full text·Original: about 4 minutes to read
In short
An InfoQ summary (October 2026) of OpenAI's DevDay 2026. The announcements: GPT-6.1 Sol, which OpenAI says approaches GPT-6 Astra on several evaluations at one-fifth of Astra's price; computer use and multi-agent features in the Agents API; Codex running in cloud environments with a code-review workflow and Codex Security Cloud; a Decisions API for picking from predefined answers; an expanded ChatGPT plugin system with MCP Events; Dots, persistent agents; and ChatGPT Space, a shared workspace for teams and agents.
Key numbers
- one-fifthGPT-6.1 Sol price relative to GPT-6 Astra's standard input and output token prices, per OpenAI
- $0.10price per million cached input tokens for GPT-6.1 Sol
As reported by the source.
English·Web page·opens infoq.com
What it reports
- GPT-6.1 Sol: an update aimed at coding, computer use and professional tasks; cached input at $0.10 per million tokens; available in the API, ChatGPT Work and Codex.
- Agents API: computer use through graphical interfaces, Codex multi-agent features, tool search, context compaction, with OpenAI running the execution infrastructure.
- Codex: cloud environments, voice input and an /agents interface in the CLI, pull-request review for GitHub and GitLab, and a security scanner that prepares fixes.
- Decisions API (limited preview): uses the smaller Luna model for classification, routing and choosing an agent's next action.
- Plugins: sidebars, interactive panels, custom file viewers, and event-triggered automations via the proposed MCP Events specification.
- Community reaction was split between interest in computer use, cloud Codex and cheaper models, and doubt that the releases go beyond existing agent platforms.
How they got it
Secondary reporting of OpenAI's announcements plus selected social-media posts. Saved by design because openai.com blocks fetchers.
Limits
- Relays vendor claims; no benchmarks, prices beyond two figures, or independent tests.
- Community quotes are anecdotal.
- Readers should check OpenAI's own release notes for details.
Why it matters for companies
The direction is the useful signal: OpenAI is moving from single model calls toward persistent agents that run on its infrastructure, operate software and share a workspace with teams. That shifts procurement questions toward where agent execution happens, what data it touches and how actions are logged. The Decisions API is a small, practical pattern worth copying with any provider: use a cheap model to route and classify against a fixed set of options.
Summary based on the full text. Read the original for the full argument and data.Read the original on infoq.com