OpenAI DevDay tracker
The keynote is over and OpenAI's official recap is live. Here are all 29 announcements, with the biggest launches first, followed by the rest grouped by product area. Each entry includes availability and a link to the official details.
Last checked September 30, 2026 at 9:15 a.m. CEST
A new top individual subscription tier for people who need the highest usage limits and OpenAI's premium speed tier. Pro 200 subscriptions are open again too, with 10 times the Plus allowance for new subscribers and no five-hour limit.
Availability: $500 per month, with 25 times the ChatGPT Plus usage allowance and Ultrafast access.
Always-on agents that learn what matters to you and handle ongoing responsibilities on your behalf. OpenAI says future options for more speed and bandwidth will be paid, while baseline functionality will remain included.
Availability: Pro and Business Premium in eligible markets. Enterprise, Edu, and Healthcare workspaces can opt into the beta. Your primary dot is included in your plan and does not draw down plan usage; Codex tasks it creates draw usage as usual.
A major Sol upgrade aimed at agentic coding, computer use, and professional work, priced at one-fifth of Astra's standard input and output token rates.
Availability: API, Plus, Pro, Business, Enterprise, and Edu.
A premium speed tier offering up to 300 tokens per second, up to 8 times faster in Codex and up to 6 times faster in the API.
Availability: GPT-6 Astra Ultrafast is available in the API and in ChatGPT Work and Codex on Pro 500 and Enterprise. GPT-6.1 Sol Ultrafast is coming soon.
Computer interaction joins the Agents API alongside multi-agent work, tool search, tool calling, and context compaction on OpenAI-managed infrastructure.
Availability: API, plus Codex and ChatGPT Work on Pro 500 and Enterprise.
Use a ChatGPT identity and plan allowance across 16 launch partners, including Devin, Notion, Vercel, T3, OpenClaw, and Dactyl, with per-tool usage controls.
Availability: Identity is global. Plan usage works for Plus and Pro users in participating tools.
A shared home where teammates, ChatGPT, and Dots work from common knowledge while ChatGPT organizes the space using team instructions.
Availability: Pro, Business, and Enterprise on desktop and web. Mobile can find, read, and share pages; mobile creation and editing are coming soon.
Run Codex locally, remotely from a phone, or in the cloud with reusable team development environments and shared permissions.
Availability: Plus, Pro, Business, Healthcare, Education, and Enterprise.
Voice control, a new /agents view for delegating and tracking parallel work, better session resumption and worktree support, and a cleaner terminal interface.
Availability: All plans.
Review summaries and diffs in the ChatGPT desktop app, ask Codex about issues, and run automatic cloud reviews for GitHub and GitLab changes.
Availability: All plans.
On-demand and scheduled repository scanning that investigates findings, removes duplicates, and prepares fixes in the cloud. Daybreak Blue model access is included.
Availability: Pro, Business, Enterprise, and Edu on desktop and web.
A Luna-powered API for choosing among finite answers using text or image context, designed for classification, routing, and agent decisions.
Availability: Limited preview today, with a broader release planned in the coming days.
OpenAI and Amazon are bringing core Agents API capabilities to a customizable managed-agent service that runs entirely inside AWS and integrates with AWS resources.
Availability: Limited preview through AWS.
Developers can give plugins a sidebar home, add interactive work panels beside conversations, and create viewers for supported file types.
Availability: All plans.
A Plugin Creator, clearer submission feedback, and better directory and conversation recommendations, with user-controlled access approvals. OpenAI says relevant plugins can be surfaced to ChatGPT's 1.2 billion weekly users.
Availability: All plans.
Sites can include supported ChatGPT plugins, use each teammate's connected data and permissions, and run shared automations.
Availability: Business, Enterprise, Healthcare, and Edu.
Support for the proposed MCP Events specification lets connected-app events trigger plugin automations, including work that continues while you are away.
Availability: All plans.
The Browser Annotation API lets sites control what people can select, attach useful context, and offer controls for previewing changes before feedback goes to ChatGPT.
Availability: Supported desktop app browsers; not currently available in Enterprise or Edu workspaces.
Collaborative documents where people and agents can write, research, create charts and images, visualize information, and gather team feedback.
Availability: Pro, Business, and Enterprise.
Create and co-edit interactive presentations with people and agents, then present in ChatGPT or export to PowerPoint or Google Slides.
Availability: Pro, Business, and Enterprise in the coming weeks.
Teams can share pages, slides, plugins, and spreadsheets, then delegate recurring or event-triggered work through connected tools.
Availability: Business and Enterprise.
Mention ChatGPT in channels, threads, or direct messages. Teammates can add context and refine the result without each person needing a ChatGPT license.
Availability: Business, Enterprise, and Edu.
Meeting notes, personalized summaries, and action items saved to ChatGPT Space. OpenAI says audio is deleted after notes are ready.
Availability: ChatGPT desktop app on macOS for Pro and Business, with Enterprise coming soon.
Profiles collect a person's activity and selected Sites in one place. Business workspace settings control whether coworkers can view workspace profiles.
Availability: Personal and Business profiles on web and desktop. Enterprise support is coming soon; Site showcases are not available on mobile.
Start a cloud-backed ChatGPT Work task on desktop, continue it on web or mobile, and let a connected computer handle steps that need local files or tools.
Availability: New synced ChatGPT Work tasks on desktop, web, and mobile; existing tasks keep their original mode.
Admins can connect company-managed app accounts so approved team workflows use shared access without requiring each participant to connect a personal account.
Availability: Eligible Team Tasks and @ChatGPT deployments in Slack and Microsoft Teams.
Admins can set workspace policies for agent tools, approvals, web access, files, sandboxing, and managed network access across supported execution environments.
Availability: Enterprise controls for local and Codex Cloud execution; ChatGPT Work Cloud has separate capability controls.
Business privacy features that combine automated safety review with stronger data protections and, later, confidential computing with verifiable controls.
Availability: Zero Data Retention with Private Safety Processing is available. Private Inference is planned for this fall.
Enterprise customers can apply part of an existing OpenAI commitment toward approved partner software. The first group includes 32 partners across creative, customer experience, legal, cybersecurity, and model infrastructure.
Availability: Eligible enterprise customers can express interest.
Published September 29, 2026
OpenAI positions GPT-6.1 Sol as near-Astra intelligence at one-fifth of Astra's standard input and output prices. It is available in the API, ChatGPT Work, and Codex, but not yet in Chat.
Vendor benchmark
OpenAI reports gains over GPT-6 Sol across its launch evaluations, along with a 95% cached-input discount. The accompanying system-card addendum documents the model's safety evaluations and safeguards.
Independent benchmark
Artificial Analysis places GPT-6.1 Sol one point below GPT-6 Astra on its Intelligence Index while reporting a cost per task below one quarter of Astra's. It also notes that Sol used more output tokens than GPT-6 Sol in its tests.
What people are building
On September 29, GitHub started rolling GPT-6.1 Sol out across Copilot, OpenAI made it the default model in Codex's bundled catalog, and Hermes Agent added direct OpenAI API and OpenRouter support. These commits prove integration work, not model quality; availability still depends on the provider and account route.
GitHub availability Codex catalog commit Hermes Agent provider commit
Run your own eval before switching.
A launch benchmark is not production proof. Compare task success, latency, total tokens, tool reliability, and cost on your own workflow before replacing a proven model.
Use Learnetto's practical AI evaluation guide and frontier model selection resources to design the comparison.
Direct posts from OpenAI and the people building these products, covering the launches drawing the most attention.
Ultrafast is available today for GPT-6 Astra in Codex, ChatGPT Work, and the API, with GPT-6.1 Sol coming soon. To access it in Codex and ChatGPT Work, we're introducing Pro 500, a new plan with our highest usage limits and access to Ultrafast.
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.
6.1 Sol at 1/5th of the price of Astra, and 95% cache read discount! Extremely capable model.
We are opening up our platform and you can now build full native apps with plugin extensions, and ship them right in ChatGPT. We have over 1.2B weekly users and will surface relevant plugins right in the conversations.