OpenAI DevDay 2026: Biggest Announcements, New AI Agents, GPT-6.1 Sol, and More
OpenAI DevDay 2026 took place on September 29, 2026, in San Francisco, bringing developers, technical founders, researchers, and AI builders together to explore new OpenAI models, APIs, agents, and developer tools. The event focused heavily on building AI systems that can do more than answer questions. For beginners, learning the AI tutorials for beginners can make these new concepts easier to understand.
The biggest announcements included OpenAI Dots, GPT-6.1 Sol, Codex Cloud improvements, the Agents API, shared ChatGPT workspaces, and new tools for businesses and developers.
Quick Summary
OpenAI DevDay 2026 focused on the next stage of AI development: autonomous agents, stronger models, coding tools, and APIs that developers can use to build complete AI-powered applications.
GPT-6.1 Sol was one of the main model announcements. OpenAI describes it as offering capabilities comparable to GPT-6 Astra while focusing on speed and affordability.
OpenAI Dots showed a different approach to AI assistants. Instead of waiting for individual prompts, Dots are designed to pursue ongoing goals across applications while using tools such as Codex and ChatGPT Work.
Codex Cloud makes software development more flexible. Developers can start and continue coding tasks from desktop, web, or mobile using isolated cloud environments.
The Agents API gives developers a way to build durable cloud agents. These agents can use tools, work inside sandboxes, edit files, execute code, and maintain sessions across tasks.
OpenAI DevDay is OpenAI’s developer-focused event for showcasing new models, APIs, tools, and technologies.
The 2026 event was held at Fort Mason in San Francisco on September 29. OpenAI described the event as a technical gathering featuring API and tool sessions, hands-on demonstrations, workshops, and opportunities for developers to connect with OpenAI teams.
The event also marked a broader shift in how OpenAI is approaching AI development.
Instead of focusing only on chatbots and traditional prompt-based applications, many of the announcements centered on AI agents that can perform multi-step tasks, coding systems that can continue working in the cloud, and APIs designed for longer-running workflows.
Biggest OpenAI DevDay 2026 Announcements
| Announcement | What it does | Main audience |
|---|---|---|
| OpenAI Dots | Always-on AI agents for ongoing tasks | Businesses and professionals |
| GPT-6.1 Sol | Faster, more affordable advanced model | Developers and businesses |
| Codex Cloud | Runs coding tasks in cloud environments | Developers |
| Agents API | Builds durable AI agents | Developers |
| ChatGPT Spaces | Shared AI workspaces | Teams |
| Developer APIs | Connect models and tools to applications | Developers |
1. OpenAI Dots: A New Type of AI Agent
One of the most notable announcements at DevDay 2026 was Dots, OpenAI’s new always-on AI agents.
According to OpenAI and reporting from Reuters, Dots are designed to pursue user goals across applications instead of simply responding to individual prompts. They can manage changing projects and use tools such as Codex and ChatGPT Work to research, analyze information, prepare documents, and build software.
For example, an AI agent could work on a project where the requirements change over time. Rather than starting from zero after every instruction, the agent can continue working toward the larger goal.
Dots can also interact through workplace platforms such as Slack and Microsoft Teams.
This represents an important change in the AI assistant model:
Traditional chatbot:
User asks → AI responds → Task ends
AI agent:
User gives goal → AI plans → Uses tools → Performs actions → Continues working → Requests approval when needed
OpenAI also highlighted controls for sensitive actions. Users can create rules for what agents can do and when they need permission. Certain sensitive actions, such as changing passwords or permanently deleting data, require explicit approval.
2. GPT-6.1 Sol
Another major announcement was GPT-6.1 Sol, the newest model in the GPT-6 family.
OpenAI describes GPT-6.1 Sol as delivering capabilities comparable to GPT-6 Astra while providing a combination of speed and affordability. OpenAI’s safety documentation also notes that the model uses the same safeguards stack as GPT-6 Astra.
OpenAI’s developer platform positions GPT-6.1 Sol for complex work and highlights it as a lower-cost option compared with its most capable models.
GPT-6.1 Sol at a glance
| Feature | GPT-6.1 Sol |
|---|---|
| Model family | GPT-6 |
| Main focus | Complex professional and agentic work |
| Developer access | OpenAI API |
| Coding | Supported |
| Computer use | Supported |
| Agent workflows | Supported |
| Positioning | Near-Astra performance at lower cost |
For developers, the importance of GPT-6.1 Sol is not simply that it is another new model.
The bigger point is model economics.
AI applications often perform multiple model calls during a workflow. Lower-cost models can make it more practical to run longer agentic processes, provided the model is capable enough for the task.
3. Codex Cloud
Codex Cloud was another important part of the DevDay-era developer updates. If you are new to AI development, our guide to AI basics for beginners is a useful starting point.
OpenAI says Codex Cloud allows users to start and continue coding tasks from desktop, web, or mobile. Each task receives an isolated workspace containing the project’s repositories, tools, and dependencies. Tasks can also continue while the user’s computer is asleep.
This changes how developers can work with AI coding agents.
A developer does not necessarily need to keep a local computer running while an AI coding task progresses in the cloud.
Why Codex Cloud matters
Codex Cloud can be useful for:
- Software development
- Code changes and refactoring
- Testing workflows
- Repository-based projects
- Longer-running coding tasks
- Team development environments
- AI-assisted engineering
OpenAI also made Codex Cloud available for enterprise environments, where authorized users can share reusable cloud environments and continue Codex tasks across devices.
4. Agents API
OpenAI’s Agents API is designed for developers who want to build durable AI agents into their own applications.
The API provides access to a managed Codex harness while OpenAI handles areas such as session management, orchestration, context compaction, and recovery. Developers can provide tools and choose the execution environment.
Agents can operate in a sandbox where they can:
- Execute code
- Edit files
- Connect to MCP servers
- Work with tools
- Produce artifacts
- Maintain durable sessions
This makes the Agents API relevant for companies that want to build more than a simple AI chatbot.
For example, a company could create an internal data analyst agent, document reviewer, incident-response agent, or software development assistant.
The important concept is that the application can give the agent an environment and tools, rather than only giving it a text prompt.
5. ChatGPT Spaces and Team Work
OpenAI also continued expanding collaborative features around ChatGPT Spaces.
OpenAI’s business updates describe shared Pages and Spaces that teams can use for projects, alongside shared plugins and recurring team tasks subject to workspace permissions.
This moves ChatGPT closer to a shared workspace rather than a purely individual assistant.
For businesses, this can help teams organize AI-assisted projects while keeping collaboration inside a controlled workspace.
6. More Focus on AI Agents
A common theme across OpenAI DevDay 2026 was the move from AI assistants to AI agents.
An assistant generally waits for instructions.
An agent can potentially:
- Receive a goal.
- Break the goal into smaller tasks.
- Use external tools.
- Execute actions.
- Check results.
- Continue working.
- Ask the user for approval when necessary.
OpenAI’s Agents API specifically supports durable sessions and managed environments, while Dots are aimed at longer-running work across applications.
This distinction matters for developers because agentic applications require more than a powerful language model. They need tools, permissions, memory or state, execution environments, monitoring, and safety controls.
OpenAI Developer Ecosystem After DevDay 2026
The broader OpenAI developer platform now includes several connected pieces. Developers who want to explore practical tools can also see our guide to the best AI tools in 2026.
| Layer | Example | Purpose |
|---|---|---|
| Models | GPT-6.1 Sol | Reasoning and generation |
| Agents | Agents API | Build autonomous workflows |
| Coding | Codex | AI-assisted software development |
| Cloud execution | Codex Cloud | Run coding tasks remotely |
| Tools | MCP and APIs | Connect external systems |
| Image generation | GPT Image 2.5 | Create and edit images |
| Collaboration | Spaces | Team-based AI work |
OpenAI’s developer platform currently highlights GPT-6.1 Sol, GPT-6 Astra Ultrafast, Codex, plugins, MCP servers, and GPT Image 2.5 as part of its broader developer ecosystem.
What OpenAI DevDay 2026 Means for Developers
For developers, the most important takeaway is that AI development is becoming increasingly workflow-oriented.
Developers can now combine models with agents, tools, cloud environments, APIs, and external services.
Instead of building an application where:
User → Prompt → Model → Answer
developers can build systems closer to:
User → Agent → Model → Tools → Data → Code → Verification → Result
That architecture can support more complex applications, including coding agents, research systems, business automation, customer support workflows, data analysis, and document processing.
However, greater autonomy also creates greater responsibility.
Agents that can access applications, files, or business systems need clear permissions and monitoring. OpenAI’s Dots announcement specifically emphasized user controls and approval requirements for sensitive actions.
OpenAI DevDay 2026 FAQs
When was OpenAI DevDay 2026?
OpenAI DevDay 2026 was held on September 29, 2026, at Fort Mason in San Francisco. OpenAI also provided a livestreamed keynote for people who could not attend in person.
What is OpenAI Dots?
Dots are OpenAI’s always-on AI agents designed to pursue goals across applications and work with tools such as Codex and ChatGPT Work.
What is GPT-6.1 Sol?
GPT-6.1 Sol is a GPT-6 family model that OpenAI positions as offering capabilities comparable to GPT-6 Astra with a focus on speed and affordability.
What is Codex Cloud?
Codex Cloud lets users start and continue coding tasks using cloud-based environments from desktop, web, or mobile. Tasks can continue even while the user’s computer is asleep.
What is the Agents API?
The Agents API lets developers build durable cloud agents using OpenAI’s managed Codex harness, tools, sessions, and execution environments.
Why was OpenAI DevDay 2026 important?
The event highlighted OpenAI’s broader move toward agentic AI, developer automation, cloud-based coding, advanced models, and applications that can perform multi-step tasks instead of only generating responses.
Final Thoughts
OpenAI DevDay 2026 was heavily focused on making AI more capable of doing real work.
Dots introduced a new approach to autonomous AI agents, GPT-6.1 Sol added another model option for complex workloads, Codex Cloud expanded AI-assisted software development, and the Agents API gave developers more infrastructure for building durable AI applications.
The bigger trend is clear: modern AI applications are moving beyond simple chat interfaces.
The next generation of AI products will increasingly combine models, agents, tools, code execution, external data, cloud environments, and human approval.
For developers and businesses, AI Era provides practical guides on where the AI ecosystem is heading, including AI app makers and other developer-focused tools.
