{"id":4642,"date":"2025-10-06T00:00:00","date_gmt":"2025-10-06T00:00:00","guid":{"rendered":"https:\/\/godofprompt.io\/blog\/2025\/10\/06\/openai-devday-2025\/"},"modified":"2025-10-06T00:00:00","modified_gmt":"2025-10-06T00:00:00","slug":"openai-devday-2025","status":"publish","type":"post","link":"https:\/\/godofprompt.ai\/blog\/openai-devday-2025\/","title":{"rendered":"OpenAI DevDay 2025: Everything Announced and What It All Means"},"content":{"rendered":"<p id>OpenAI held their annual developer conference on October 7, 2025, in San Francisco. <\/p>\n<p id>They announced a lot of new products, tools, and features in one morning.<\/p>\n<p id>Let me walk you through every announcement, what each one does, and why you should care about it.<\/p>\n<p id><strong id>ALSO&nbsp;READ:<\/strong> <a href=\"https:\/\/godofprompt.ai\/blog\/ultimate-guide-to-openai-sora-2\" id>Ultimate Guide to OpenAI Sora 2: Everything You Need to Know in 2025<\/a><\/p>\n<figure id class=\"w-richtext-figure-type-image w-richtext-align-fullwidth\" style=\"max-width:1200px\" data-rt-type=\"image\" data-rt-align=\"fullwidth\" data-rt-max-width=\"1200px\"><a id><\/p>\n<div id><img decoding=\"async\" src=\"https:\/\/godofprompt.ai\/blog\/wp-content\/uploads\/2026\/05\/69ea6cba6c0e633fc8d270ff_6956a4d5674c51adc7dce1e8_675f5a351b3337145eb8c021_BiggestAIPromptLibrary_OpenGraph_Button.webp\" loading=\"lazy\" alt=\"__wf_reserved_inherit\" width=\"auto\" height=\"auto\" id><\/div>\n<p><\/a><figcaption id>Discover The <a href=\"https:\/\/godofprompt.ai\/prompt-library\" id>Biggest AI Prompt Library<\/a> by God Of Prompt<\/figcaption><\/figure>\n<h2 id>The Big Picture: How Much Things Changed<\/h2>\n<p id>Before we get into the new stuff, here&#8217;s how much the platform grew since the last DevDay in 2023.<\/p>\n<p id><strong id>Two years ago at DevDay 2023:<\/strong><\/p>\n<ul id>\n<li id>2 million developers building with OpenAI<\/li>\n<li id>100 million people using ChatGPT each week<\/li>\n<li id>300 million tokens running through the API every minute<\/li>\n<\/ul>\n<p id><strong id>Now at DevDay 2025:<\/strong><\/p>\n<ul id>\n<li id>4 million developers on the platform<\/li>\n<li id>800 million people using ChatGPT each week<\/li>\n<li id>6 billion tokens per minute through the API<\/li>\n<\/ul>\n<figure id class=\"w-richtext-figure-type-image w-richtext-align-fullwidth\" style=\"max-width:1408px\" data-rt-type=\"image\" data-rt-align=\"fullwidth\" data-rt-max-width=\"1408px\">\n<div id><img decoding=\"async\" src=\"https:\/\/godofprompt.ai\/blog\/wp-content\/uploads\/2026\/05\/69ea6cba6c0e633fc8d277f8_6956a662fef75578c38359f4_68e45bb3130ca26a93ced1d4_The-Big-Picture-How-M.avif\" loading=\"lazy\" alt=\"The Big Picture How Much Things Changed\" width=\"auto\" height=\"auto\" id><\/div><figcaption id>The Big Picture How Much Things Changed<\/figcaption><\/figure>\n<p id>The developer count doubled. <\/p>\n<p id>The user count went up 8 times. <\/p>\n<p id>The token processing increased 20 times.<\/p>\n<p id>This matters because all the announcements are built for this bigger scale. <\/p>\n<p id>OpenAI isn&#8217;t making experimental features anymore. <\/p>\n<p id>They&#8217;re building for production use by millions of developers.<\/p>\n<h3 id>Announcement #1: Apps You Can Build Inside ChatGPT<\/h3>\n<p id>OpenAI announced that developers can now build full applications that work inside ChatGPT conversations.<\/p>\n<p id>This isn&#8217;t like the GPTs they released before. <\/p>\n<p id>This is much bigger.<\/p>\n<h3 id>What Apps in ChatGPT Actually Are<\/h3>\n<p id>Apps are interactive programs that live right in the chat. <\/p>\n<p id>Users don&#8217;t download anything. <\/p>\n<p id>They don&#8217;t visit another website. <\/p>\n<p id>The app runs inside the conversation.<\/p>\n<p id><strong id>How users find apps:<\/strong><\/p>\n<p id>They can call them by name. <\/p>\n<p id>Someone types <\/p>\n<blockquote id><p>&#8220;Spotify, make me a workout playlist&#8221; <\/p><\/blockquote>\n<p id>and the Spotify app appears in the chat. Or ChatGPT suggests them. <\/p>\n<p id>If you&#8217;re talking about buying a house, ChatGPT might suggest the Zillow app to browse homes.<\/p>\n<p id><strong id>What makes these different from chatbots:<\/strong><\/p>\n<p id>These apps have real interfaces. <\/p>\n<p id>Maps. Video players. Interactive forms. Clickable buttons. You can do real work inside them.<\/p>\n<p id>They connect to actual backends. Your databases. Your APIs. Your business logic. <\/p>\n<p id>This isn&#8217;t just AI responses &#8211; it&#8217;s your full application.<\/p>\n<p id>Users can log in with existing accounts. If someone already pays for your service, they can access it through ChatGPT.<\/p>\n<h3 id>The Seven Apps Available Today<\/h3>\n<figure id class=\"w-richtext-figure-type-image w-richtext-align-fullwidth\" style=\"max-width:1024px\" data-rt-type=\"image\" data-rt-align=\"fullwidth\" data-rt-max-width=\"1024px\">\n<div id><img decoding=\"async\" src=\"https:\/\/godofprompt.ai\/blog\/wp-content\/uploads\/2026\/05\/69ea6cba6c0e633fc8d277ef_6956a662fef75578c3835a10_68e45bed1d5a896651d0097a_The-Seven-Apps-Available-25.avif\" loading=\"lazy\" alt=\"The Seven Apps Available Today\" width=\"auto\" height=\"auto\" id><\/div><figcaption id>The Seven Apps Available Today<\/figcaption><\/figure>\n<p id>OpenAI launched with seven partner companies. Each one shows what&#8217;s possible.<\/p>\n<p id><strong id>Booking.com<\/strong> &#8211; Search for hotels, see results with photos and prices, filter by what you need, book directly<\/p>\n<p id><strong id>Canva<\/strong> &#8211; Type an outline, Canva turns it into a slide deck. Edit the design right there. Export when you&#8217;re done.<\/p>\n<p id><strong id>Coursera<\/strong> &#8211; Watch educational videos with AI explaining difficult parts as you watch. Ask questions about what you&#8217;re learning.<\/p>\n<p id><strong id>Expedia<\/strong> &#8211; Plan trips, search flights, book hotels and activities, all in the conversation<\/p>\n<p id><strong id>Figma<\/strong> &#8211; Upload a sketch, Figma converts it to a working diagram. Click to open in FigJam for more editing.<\/p>\n<p id><strong id>Spotify<\/strong> &#8211; Describe what kind of music you want, Spotify builds a playlist. Refine it by talking.<\/p>\n<p id><strong id>Zillow<\/strong> &#8211; See homes for sale on an interactive map. Filter by price, bedrooms, features. Ask about neighborhoods.<\/p>\n<h3 id>The Apps SDK: How Developers Build These<\/h3>\n<p id>OpenAI released the Apps SDK so any developer can build apps for ChatGPT.<\/p>\n<p id><strong id>What the SDK gives you:<\/strong><\/p>\n<p id>It&#8217;s built on something called Model Context Protocol, or MCP. <\/p>\n<p id>This is an open standard for connecting AI to tools. <\/p>\n<p id>If you already know MCP, you&#8217;re halfway there.<\/p>\n<p id>The Apps SDK adds more on top of MCP:<\/p>\n<ul id>\n<li id>Ways to create user interfaces<\/li>\n<li id>Ways to connect to your backend servers<\/li>\n<li id>Ways to handle user login<\/li>\n<li id>Ways to manage what&#8217;s happening in the conversation<\/li>\n<\/ul>\n<p id><strong id>What you can build:<\/strong><\/p>\n<figure id class=\"w-richtext-figure-type-image w-richtext-align-fullwidth\" style=\"max-width:1248px\" data-rt-type=\"image\" data-rt-align=\"fullwidth\" data-rt-max-width=\"1248px\">\n<div id><img decoding=\"async\" src=\"https:\/\/godofprompt.ai\/blog\/wp-content\/uploads\/2026\/05\/69ea6cba6c0e633fc8d277f6_6956a662fef75578c38359fa_68e45c2b79fc76ab77f9089d_What-you-can-build.avif\" loading=\"lazy\" alt=\"__wf_reserved_inherit\" width=\"auto\" height=\"auto\" id><\/div>\n<\/figure>\n<p id>Your app can use any web technology. <\/p>\n<p id>React, Vue, plain HTML and CSS &#8211; whatever you know.<\/p>\n<p id>You control the backend completely. <\/p>\n<p id>Your servers, your database, your security, your business rules.<\/p>\n<p id>You can charge money. <\/p>\n<p id>Let free users see some features, paid users see everything. You decide.<\/p>\n<p id><strong id>The &#8220;talking to apps&#8221; feature:<\/strong><\/p>\n<p id>This is important. <\/p>\n<p id>Your app can tell ChatGPT what the user is looking at.<\/p>\n<p id><strong id>Example:<\/strong> Someone is watching a course in the Coursera app. <\/p>\n<blockquote id><p>They ask &#8220;can you explain this part more?&#8221; <\/p><\/blockquote>\n<p id>ChatGPT knows exactly which part of the video they mean because the Coursera app told it.<\/p>\n<p id>Or someone is looking at a house on the Zillow map. <\/p>\n<blockquote id><p>They ask &#8220;how close is this to a dog park?&#8221; <\/p><\/blockquote>\n<p id>ChatGPT knows which house they mean.<\/p>\n<p id>This makes the whole thing feel natural. <\/p>\n<p id>No explaining context over and over.<\/p>\n<p id><strong id>When you can publish your app:<\/strong><\/p>\n<p id>Right now, the SDK is in preview. <\/p>\n<p id>You can build and test using something called Developer Mode.<\/p>\n<p id>Later in 2025, OpenAI will open submissions. <\/p>\n<p id>Any developer can submit their app for review.<\/p>\n<p id>When approved, your app goes in a directory where people can find it.<\/p>\n<p id><strong id>How you make money:<\/strong><\/p>\n<p id>OpenAI hasn&#8217;t released all the details yet, but they said it&#8217;s coming later this year.<\/p>\n<p id>One option they mentioned: Agentic Commerce Protocol. <\/p>\n<p id>This lets people buy things instantly inside ChatGPT without leaving to another website.<\/p>\n<h3 id>Why This Matters So Much<\/h3>\n<p id>Getting users is the hardest part of building software. <\/p>\n<p id>You can build something great and nobody finds it.<\/p>\n<p id>Apps SDK solves this. <\/p>\n<p id>Build your app once. It reaches 800 million people who use ChatGPT every week. <\/p>\n<p id>They find it naturally when they need it.<\/p>\n<p id>No paying for ads. No fighting with app store policies. <\/p>\n<p id>No long download and setup process.<\/p>\n<p id>That&#8217;s a massive change in how software distribution works.<\/p>\n<h3 id>Announcement #2: AgentKit &#8211; The Complete Platform for Building Agents<\/h3>\n<p id>OpenAI announced AgentKit, which they describe as a complete set of tools for building AI agents.<\/p>\n<h3 id>The Problem AgentKit Solves<\/h3>\n<figure id class=\"w-richtext-figure-type-image w-richtext-align-fullwidth\" style=\"max-width:1024px\" data-rt-type=\"image\" data-rt-align=\"fullwidth\" data-rt-max-width=\"1024px\">\n<div id><img decoding=\"async\" src=\"https:\/\/godofprompt.ai\/blog\/wp-content\/uploads\/2026\/05\/69ea6cba6c0e633fc8d277ea_6956a662fef75578c3835a08_68e45cb279fc76ab77f91fa0_The-Problem-AgentKit-Solves.avif\" loading=\"lazy\" alt=\"__wf_reserved_inherit\" width=\"auto\" height=\"auto\" id><\/div><figcaption id>The Problem AgentKit Solves<\/figcaption><\/figure>\n<p id>Right now, building an AI agent that actually works in production is really hard.<\/p>\n<p id>You need to connect a bunch of different tools. Set up complicated workflows. <\/p>\n<p id>Write custom code to connect everything. Build evaluation systems to test if it works. Create a user interface. Handle errors. Make it secure.<\/p>\n<p id>Most agents never make it past the prototype stage because it&#8217;s too much work.<\/p>\n<p id>AgentKit changes this. It gives you everything in one place.<\/p>\n<h3 id>The Four Parts of AgentKit<\/h3>\n<p id>AgentKit has four main components that work together.<\/p>\n<h4 id>Part 1: Agent Builder<\/h4>\n<p id>This is a visual tool for designing agent workflows.<\/p>\n<p id>Instead of writing code, you drag boxes onto a canvas and connect them. <\/p>\n<p id>Each box is a different step in what your agent does.<\/p>\n<h3 id><strong id>Types of boxes you can use:<\/strong><\/h3>\n<p id><strong id>Agent boxes<\/strong> &#8211; These are AI agents for specific tasks. You write instructions for what each one does, give it tools to use, connect it to your data.<\/p>\n<p id><strong id>Tool boxes<\/strong> &#8211; These connect to file search, APIs, databases, or MCP servers. Any external thing your agent needs to access.<\/p>\n<p id><strong id>Guardrail boxes<\/strong> &#8211; These are safety filters. They can block personal information, catch attempts to trick the agent, or filter inappropriate content.<\/p>\n<p id><strong id>Logic boxes<\/strong> &#8211; These handle if\/then decisions. &#8220;If the user asks about X, go to agent A. If they ask about Y, go to agent B.&#8221;<\/p>\n<p id><strong id>Approval boxes<\/strong> &#8211; These stop and ask a human before doing something. Good for sensitive actions like spending money or sharing data.<\/p>\n<p id><strong id>Widget boxes<\/strong> &#8211; These create custom response formats. Like showing data in a table, or displaying a map, or creating an interactive form.<\/p>\n<h3 id><strong id>How you use it:<\/strong><\/h3>\n<figure id class=\"w-richtext-figure-type-image w-richtext-align-fullwidth\" style=\"max-width:1024px\" data-rt-type=\"image\" data-rt-align=\"fullwidth\" data-rt-max-width=\"1024px\">\n<div id><img decoding=\"async\" src=\"https:\/\/godofprompt.ai\/blog\/wp-content\/uploads\/2026\/05\/69ea6cba6c0e633fc8d277eb_6956a662fef75578c38359ee_68e45d009dc0c55681c20354_how-to-use-it-agentki.avif\" loading=\"lazy\" alt=\"__wf_reserved_inherit\" width=\"auto\" height=\"auto\" id><\/div><figcaption id><strong id>How you use it<\/strong><\/figcaption><\/figure>\n<p id>You start with a blank canvas. <\/p>\n<p id>Add the boxes you need. <\/p>\n<p id>Connect them in order. <\/p>\n<p id>Click on each box to configure what it does. <\/p>\n<p id>Run a test conversation to see if it works. <\/p>\n<p id>Adjust anything that&#8217;s wrong. <\/p>\n<p id>Run it again. <\/p>\n<p id>Keep going until it works right.<\/p>\n<p id><strong id>Real example from the conference:<\/strong><\/p>\n<p id>Someone built an agent live on stage in under 8 minutes. <\/p>\n<p id>It could answer questions about the DevDay schedule, show session details in a nice format, and respond in a specific style. <\/p>\n<p id>All using Agent Builder.<\/p>\n<p id>Companies like Ramp said Agent Builder let them build in hours what used to take months.<\/p>\n<p id><strong id>When to use code instead:<\/strong><\/p>\n<p id>You can export your workflow to code if you need to customize something the visual tool can&#8217;t do. <\/p>\n<p id>Or if you want to integrate it into existing code.<\/p>\n<p id>The workflow gets an ID either way, so you can switch between visual and code approaches.<\/p>\n<h4 id>Part 2: Connector Registry<\/h4>\n<p id>This is an admin panel for managing all your data connections in one place.<\/p>\n<p id><strong id>What it connects:<\/strong><\/p>\n<ul id>\n<li id>Google Drive<\/li>\n<li id>Dropbox<\/li>\n<li id>SharePoint<\/li>\n<li id>Microsoft Teams<\/li>\n<li id>OneDrive<\/li>\n<li id>Any MCP server<\/li>\n<\/ul>\n<p id><strong id>Why this matters:<\/strong><\/p>\n<p id>Without Connector Registry, every team rebuilds the same connections. <\/p>\n<p id>Your support team connects to Google Drive. <\/p>\n<p id>Your sales team connects to Google Drive. <\/p>\n<p id>Your product team connects to Google Drive. <\/p>\n<p id>Three times, three different ways.<\/p>\n<p id>With Connector Registry, you connect once. <\/p>\n<p id>It&#8217;s available everywhere across your company&#8217;s ChatGPT and API usage.<\/p>\n<p id><strong id>What admins can control:<\/strong><\/p>\n<ul id>\n<li id>Who is allowed to access which data sources<\/li>\n<li id>Which teams or workspaces get which connectors<\/li>\n<li id>Security settings for each connection<\/li>\n<li id>Monitoring who uses what<\/li>\n<\/ul>\n<p id>For big companies with lots of teams and lots of data, this saves a huge amount of time and keeps things secure.<\/p>\n<h4 id>Part 3: ChatKit<\/h4>\n<figure id class=\"w-richtext-figure-type-image w-richtext-align-fullwidth\" style=\"max-width:1024px\" data-rt-type=\"image\" data-rt-align=\"fullwidth\" data-rt-max-width=\"1024px\">\n<div id><img decoding=\"async\" src=\"https:\/\/godofprompt.ai\/blog\/wp-content\/uploads\/2026\/05\/69ea6cba6c0e633fc8d277f1_6956a662fef75578c38359f1_68e45d858da7646d506e0f94_Part-3-ChatKit.avif\" loading=\"lazy\" alt=\"__wf_reserved_inherit\" width=\"auto\" height=\"auto\" id><\/div><figcaption id>Part 3: ChatKit<\/figcaption><\/figure>\n<p id>ChatKit is a pre-built chat interface you can drop into your app.<\/p>\n<p id><strong id>What it handles for you:<\/strong><\/p>\n<p id>Building a good chat interface is harder than it looks. <\/p>\n<p id>You need to handle messages streaming in character by character. Save conversation history. <\/p>\n<p id>Show when the AI is thinking. Display rich content like images or interactive widgets. <\/p>\n<p id>Work on phones and computers.<\/p>\n<p id>ChatKit does all of this. You just add it to your app and customize the colors and branding.<\/p>\n<p id><strong id>How fast it is to use:<\/strong><\/p>\n<p id>Canva said they built a support agent for their developer community and got it working in under an hour using ChatKit. <\/p>\n<p id>That saved them two weeks compared to building it themselves.<\/p>\n<p id><strong id>What you customize:<\/strong><\/p>\n<ul id>\n<li id>Colors to match your brand<\/li>\n<li id>Font and text sizes<\/li>\n<li id>What the chat says when it starts<\/li>\n<li id>The placeholder text in the input box<\/li>\n<li id>Any custom buttons or actions<\/li>\n<\/ul>\n<p id>It works with agents from Agent Builder, or with custom agents you build yourself.<\/p>\n<h4 id>Part 4: Evaluation Tools<\/h4>\n<p id>You can&#8217;t make agents better if you can&#8217;t measure how they&#8217;re doing. AgentKit includes tools for this.<\/p>\n<p id><strong id>Trace grading<\/strong> &#8211; This shows you every decision the agent made, step by step. &#8220;It classified the request as type A. <\/p>\n<p id>It routed to agent 1. Agent 1 called tool X. Tool X returned data Y.&#8221; You see the whole chain.<\/p>\n<p id><strong id>Datasets<\/strong> &#8211; You create test cases. The same questions or scenarios over and over. Run your agent against them. <\/p>\n<p id>See where it succeeds and where it fails. This helps you find and fix problems.<\/p>\n<p id><strong id>Automated prompt optimization<\/strong> &#8211; The system tries different ways of writing instructions to your agents. It finds which versions work best. This saves you from manually testing dozens of variations.<\/p>\n<p id><strong id>External model testing<\/strong> &#8211; You can test your agent using other AI models from other companies, not just OpenAI. <\/p>\n<p id>This helps you compare and choose the best one for your use case.<\/p>\n<h3 id>Guardrails: The Safety Layer<\/h3>\n<p id>Guardrails are part of AgentKit but important enough to explain separately.<\/p>\n<p id><strong id>What they are:<\/strong><\/p>\n<p id>Safety filters that check what goes into and comes out of your agents.<\/p>\n<p id><strong id>Types of guardrails:<\/strong><\/p>\n<p id><strong id>PII masking<\/strong> &#8211; Automatically finds and hides personal information. <\/p>\n<p id>Names, emails, phone numbers, social security numbers, credit card numbers. You define what counts as sensitive.<\/p>\n<p id><strong id>Jailbreak detection<\/strong> &#8211; Catches attempts to trick your agent into doing things it shouldn&#8217;t. <\/p>\n<p id>People try creative ways to bypass rules. Guardrails block these attempts.<\/p>\n<p id><strong id>Content moderation<\/strong> &#8211; Filters inappropriate content, harmful instructions, or things that violate your policies.<\/p>\n<p id><strong id>Custom rules<\/strong> &#8211; You can write your own. Check for anything specific to your business or use case.<\/p>\n<h3 id><strong id>How to use them:<\/strong><\/h3>\n<p id>In Agent Builder, you drag a guardrail box into your workflow. Put it before sensitive steps. Configure what it should catch. That&#8217;s it.<\/p>\n<p id>Or use them as code libraries in Python or JavaScript. Add a few lines to check inputs or outputs.<\/p>\n<p id><strong id>Performance:<\/strong><\/p>\n<p id>Guardrails are fast. Most add 10 to 50 milliseconds. That&#8217;s tiny compared to how long AI models take to think.<\/p>\n<p id><strong id>Open source:<\/strong><\/p>\n<p id>OpenAI made Guardrails open source. You can see exactly how they work. Contribute improvements. Trust them because you can inspect the code.<\/p>\n<h3 id>Real Companies Using AgentKit<\/h3>\n<figure id class=\"w-richtext-figure-type-image w-richtext-align-fullwidth\" style=\"max-width:1024px\" data-rt-type=\"image\" data-rt-align=\"fullwidth\" data-rt-max-width=\"1024px\">\n<div id><img decoding=\"async\" src=\"https:\/\/godofprompt.ai\/blog\/wp-content\/uploads\/2026\/05\/69ea6cba6c0e633fc8d277ed_6956a662fef75578c3835a05_68e45dbb3a516cfa32e58d0f_Real-Companies-Using-AgentK.avif\" loading=\"lazy\" alt=\"__wf_reserved_inherit\" width=\"auto\" height=\"auto\" id><\/div><figcaption id>Real Companies Using AgentKit<\/figcaption><\/figure>\n<p id><strong id>Albertsons<\/strong> &#8211; They run over 2,000 grocery stores. <\/p>\n<p id>Store managers have to make constant decisions about promotions, inventory, displays. <\/p>\n<p id>They built an agent with AgentKit. <\/p>\n<p id>Now managers ask questions like &#8220;Why are ice cream sales down 32%?&#8221; <\/p>\n<p id>The agent analyzes everything and suggests what to do.<\/p>\n<p id><strong id>HubSpot<\/strong> &#8211; They built an agent called Breeze using AgentKit. <\/p>\n<p id>It answers customer support questions by searching their knowledge base, pulling policy details, and putting together smart answers. <\/p>\n<p id>Faster help for customers.<\/p>\n<p id><strong id>Ramp<\/strong> &#8211; They said Agent Builder cut their development time by 70%. <\/p>\n<p id>What used to take two quarters now takes two sprints. <\/p>\n<p id>The visual canvas kept their legal team, product team, and engineering team all aligned.<\/p>\n<p id><strong id>LY Corporation<\/strong> &#8211; A big technology company in Japan. <\/p>\n<p id>They built a work assistant agent in under two hours using Agent Builder. <\/p>\n<p id>Engineers and business people worked together in the same interface.<\/p>\n<h3 id>Who Can Use AgentKit<\/h3>\n<p id>Available now if you have:<\/p>\n<ul id>\n<li id>ChatGPT Plus (personal plan)<\/li>\n<li id>ChatGPT Pro (better personal plan)<\/li>\n<li id>Business plan<\/li>\n<li id>Education plan<\/li>\n<li id>Enterprise plan<\/li>\n<\/ul>\n<h3 id>Announcement #3: Codex Is Now Officially Released<\/h3>\n<p id>Codex has been in <strong id>&#8220;research preview&#8221;<\/strong> since earlier this year. <\/p>\n<p id>Now it&#8217;s officially released and generally available. <\/p>\n<p id>Plus three big new features.<\/p>\n<h3 id>What Codex Is<\/h3>\n<p id>Codex is an AI agent that helps you write code.<\/p>\n<p id>It works in several places:<\/p>\n<ul id>\n<li id>Your code editor (VS Code, other IDEs)<\/li>\n<li id>Your terminal (command line)<\/li>\n<li id>GitHub (for reviewing pull requests)<\/li>\n<li id>The cloud (for remote development)<\/li>\n<\/ul>\n<p id>Everything connects through your ChatGPT account. You can start work in one place and continue in another.<\/p>\n<h3 id>How Much It&#8217;s Being Used<\/h3>\n<p id>OpenAI shared some numbers:<\/p>\n<p id>Since early August, daily messages to Codex went up 10 times. <\/p>\n<p id>People are using it way more.<\/p>\n<p id>The model powering it, GPT-5-Codex, served over 40 trillion tokens in just three weeks. <\/p>\n<p id>That&#8217;s a massive amount.<\/p>\n<p id>Inside OpenAI, almost every engineer uses Codex now. <\/p>\n<p id>They merge 70% more pull requests each week compared to before. <\/p>\n<p id>And Codex automatically reviews almost every pull request to catch problems.<\/p>\n<h3 id>The Three New Features<\/h3>\n<figure id class=\"w-richtext-figure-type-image w-richtext-align-fullwidth\" style=\"max-width:1408px\" data-rt-type=\"image\" data-rt-align=\"fullwidth\" data-rt-max-width=\"1408px\">\n<div id><img decoding=\"async\" src=\"https:\/\/godofprompt.ai\/blog\/wp-content\/uploads\/2026\/05\/69ea6cba6c0e633fc8d277f4_6956a662fef75578c38359f7_68e45dfc2e5115e7071cd0c4_The-Three-New-Features.avif\" loading=\"lazy\" alt=\"__wf_reserved_inherit\" width=\"auto\" height=\"auto\" id><\/div>\n<\/figure>\n<h4 id>New Feature 1: Slack Integration<\/h4>\n<p id>You can now use Codex directly in Slack.<\/p>\n<p id><strong id>How it works:<\/strong><\/p>\n<p id>In any Slack channel or thread, tag @Codex and ask it to do something.<\/p>\n<blockquote id><p>&#8220;@Codex add error handling to the authentication endpoint&#8221;<\/p><\/blockquote>\n<p id>Codex reads the whole thread to understand context. <\/p>\n<p id>It figures out which code you&#8217;re talking about. <\/p>\n<p id>Makes the changes. Posts a link where you can review what it did.<\/p>\n<p id><strong id>What you can do with the results:<\/strong><\/p>\n<p id>Click the link to see the changes in Codex Cloud. <\/p>\n<p id>From there you can merge the changes to your code, keep asking for more changes, or download it to work on your own computer.<\/p>\n<p id><strong id>Why this matters:<\/strong><\/p>\n<p id>Your team already works in Slack. <\/p>\n<p id>Now coding help is right there too. <\/p>\n<p id>No switching to another tool. <\/p>\n<p id>Just ask Codex like you&#8217;d ask a teammate.<\/p>\n<h4 id>New Feature 2: Codex SDK<\/h4>\n<p id>The SDK (Software Development Kit) is a way to use Codex in your own tools and automated workflows.<\/p>\n<p id><strong id>What it&#8217;s for:<\/strong><\/p>\n<ul id>\n<li id>Automatically running Codex in your build pipeline<\/li>\n<li id>Creating custom development tools<\/li>\n<li id>Connecting Codex to GitHub Actions<\/li>\n<li id>Using Codex in scripts<\/li>\n<\/ul>\n<p id><strong id>Example code:<\/strong><\/p>\n<blockquote id><p><code id>import { Codex } from \"@openai\/codex-sdk\";<\/p>\n<p>const agent = new Codex();<br \/>const thread = await agent.startThread();<br \/>const result = await thread.run(\"Refactor this function\");<br \/><\/code><\/p><\/blockquote>\n<p id>That&#8217;s TypeScript. Support for more programming languages is coming.<\/p>\n<p id><strong id>Real use case:<\/strong><\/p>\n<p id>You can set up GitHub Actions to automatically review every pull request with Codex. <\/p>\n<p id>It checks for security issues, suggests improvements, makes sure the code follows your standards. <\/p>\n<p id>All automatic.<\/p>\n<h4 id>New Feature 3: Admin Tools for Companies<\/h4>\n<p id>If you&#8217;re a company using Codex, admins now have more control.<\/p>\n<p id><strong id>What admins can do:<\/strong><\/p>\n<ul id>\n<li id>Edit or delete Codex Cloud environments (the remote coding spaces)<\/li>\n<li id>Set configuration rules that everyone has to follow<\/li>\n<li id>Monitor what actions Codex takes<\/li>\n<li id>See usage statistics across the whole company<\/li>\n<li id>View analytics about code review quality<\/li>\n<\/ul>\n<p id><strong id>Example analytics:<\/strong><\/p>\n<p id>Dashboards show things like &#8220;daily code review issues by priority&#8221; and &#8220;sentiment of code review feedback over time.&#8221;<\/p>\n<p id>This helps managers understand if Codex is actually helping and where teams might need training.<\/p>\n<h3 id>The Model Behind It: GPT-5-Codex<\/h3>\n<p id>Codex now uses a model called GPT-5-Codex. <\/p>\n<p id>OpenAI trained this specifically for coding tasks.<\/p>\n<p id><strong id>What makes it special:<\/strong><\/p>\n<p id>It&#8217;s better at refactoring code (reorganizing and improving existing code). <\/p>\n<p id>It&#8217;s better at reviewing code. <\/p>\n<p id>And it can think for different amounts of time depending on how hard the problem is. <\/p>\n<p id>Simple problems get quick answers. <\/p>\n<p id>Hard problems get more thinking time.<\/p>\n<h3 id>Companies Using It<\/h3>\n<p id><strong id>Cisco<\/strong> &#8211; They rolled out Codex to their entire engineering organization. <\/p>\n<p id>Code reviews that used to take a long time now happen 50% faster. Projects that took weeks now take days.<\/p>\n<p id><strong id>Instacart<\/strong> &#8211; They connected the Codex SDK to their internal system called Olive. <\/p>\n<p id>Now Codex automatically cleans up old code nobody uses anymore, removes expired experiments, and improves performance. <\/p>\n<p id>Engineers focus on new features instead of cleanup.<\/p>\n<h3 id>Pricing Change<\/h3>\n<p id>Starting October 20, when you use Codex Cloud (the remote version), it counts against your usage limits. <\/p>\n<p id>Before this, it was separate. <\/p>\n<p id>Check the pricing page for your specific plan to see what this means for you.<\/p>\n<h3 id>Announcement #4: GPT-5 Pro in the API<\/h3>\n<figure id class=\"w-richtext-figure-type-image w-richtext-align-fullwidth\" style=\"max-width:1408px\" data-rt-type=\"image\" data-rt-align=\"fullwidth\" data-rt-max-width=\"1408px\">\n<div id><img decoding=\"async\" src=\"https:\/\/godofprompt.ai\/blog\/wp-content\/uploads\/2026\/05\/69ea6cba6c0e633fc8d277ec_6956a662fef75578c38359fd_68e45e3d0e4ac7ada20b2795_Announcement-2525234-GPT-5-25252.avif\" loading=\"lazy\" alt=\"__wf_reserved_inherit\" width=\"auto\" height=\"auto\" id><\/div><figcaption id>Announcement #4: GPT-5 Pro in the API<\/figcaption><\/figure>\n<p id>OpenAI released GPT-5 a while ago. <\/p>\n<p id>Now they released GPT-5 Pro to all developers through the API.<\/p>\n<h3 id>What GPT-5 Pro Is<\/h3>\n<p id>It&#8217;s OpenAI&#8217;s smartest, most capable model. B<\/p>\n<p id>uilt for tasks where you absolutely need accuracy and can&#8217;t afford mistakes.<\/p>\n<h3 id>When You Should Use It<\/h3>\n<p id><strong id>Use GPT-5 Pro for:<\/strong><\/p>\n<ul id>\n<li id>Financial analysis where numbers have to be exactly right<\/li>\n<li id>Legal document review where missing something could be serious<\/li>\n<li id>Healthcare applications where wrong information could hurt someone<\/li>\n<li id>Any complex reasoning where accuracy matters more than speed or cost<\/li>\n<\/ul>\n<p id><strong id>Use regular GPT-5 or GPT-4 for:<\/strong><\/p>\n<ul id>\n<li id>Things where speed is more important than perfection<\/li>\n<li id>Simple, straightforward tasks<\/li>\n<li id>Processing huge amounts of stuff where cost adds up<\/li>\n<li id>Situations where an occasional small mistake is okay<\/li>\n<\/ul>\n<h3 id>How It&#8217;s Different<\/h3>\n<p id>GPT-5 Pro thinks longer about problems. <\/p>\n<p id>It shows you its reasoning process. <\/p>\n<p id>It catches its own mistakes. <\/p>\n<p id>This all costs more and takes more time, but you get better answers.<\/p>\n<h3 id>Availability<\/h3>\n<p id>It&#8217;s in the API right now. All developers can use it.<\/p>\n<h3 id>Announcement #5: Sora 2 Video API<\/h3>\n<figure id class=\"w-richtext-figure-type-image w-richtext-align-fullwidth\" style=\"max-width:1408px\" data-rt-type=\"image\" data-rt-align=\"fullwidth\" data-rt-max-width=\"1408px\">\n<div id><img decoding=\"async\" src=\"https:\/\/godofprompt.ai\/blog\/wp-content\/uploads\/2026\/05\/69ea6cba6c0e633fc8d27817_6956a662fef75578c3835a02_68e45e563a516cfa32e5d4b4_Announcement-2525235-Sora-.avif\" loading=\"lazy\" alt=\"__wf_reserved_inherit\" width=\"auto\" height=\"auto\" id><\/div><figcaption id>Announcement #5 Sora 2 Video API<\/figcaption><\/figure>\n<p id>Sora 2 is OpenAI&#8217;s video generation model. <\/p>\n<p id>It&#8217;s been available in the Sora app. <\/p>\n<p id>Now there&#8217;s an API so developers can generate videos programmatically.<\/p>\n<h3 id>Two Versions of the Model<\/h3>\n<p id><strong id>sora-2<\/strong> &#8211; The faster one<\/p>\n<p id>Good quality results but generates quickly. Perfect for when you&#8217;re testing ideas and iterating. Costs less.<\/p>\n<p id><strong id>sora-2-pro<\/strong> &#8211; The quality one<\/p>\n<p id>Takes longer to generate. Costs more. But the quality is higher and more polished. Use this for final production videos.<\/p>\n<h3 id>How the API Works<\/h3>\n<p id><strong id>Five different endpoints:<\/strong><\/p>\n<ol id>\n<li id><strong id>Create<\/strong> &#8211; Start making a video by sending a text description<\/li>\n<li id><strong id>Get status<\/strong> &#8211; Check if your video is done yet<\/li>\n<li id><strong id>Download<\/strong> &#8211; Get the finished MP4 file<\/li>\n<li id><strong id>List<\/strong> &#8211; See all your videos<\/li>\n<li id><strong id>Delete<\/strong> &#8211; Remove videos you don&#8217;t need<\/li>\n<\/ol>\n<p id><strong id>The process:<\/strong><\/p>\n<p id>Generating video takes time. You send your request. <\/p>\n<p id>The API gives you an ID. You check the status until it says &#8220;completed.&#8221; Then you download it.<\/p>\n<p id>Or instead of checking repeatedly, you can set up webhooks. OpenAI notifies you when it&#8217;s done.<\/p>\n<h3 id>What Videos Include<\/h3>\n<p id>Every video comes with audio. <\/p>\n<p id>Not just any audio &#8211; synchronized audio that matches what&#8217;s happening. <\/p>\n<p id>Dialogue, sound effects, background noise, all fitting together.<\/p>\n<h3 id>Advanced Options<\/h3>\n<p id><strong id>Image references<\/strong> &#8211; You can upload an image as the starting frame. <\/p>\n<p id>Sora 2 animates from there. Useful for brand consistency or specific characters.<\/p>\n<p id><strong id>Remix<\/strong> &#8211; Take an existing video and change specific things. &#8220;Make the sky purple&#8221; or &#8220;add a second character.&#8221; <\/p>\n<p id>It modifies instead of starting over.<\/p>\n<p id><strong id>Parameters you control:<\/strong><\/p>\n<ul id>\n<li id>How long the video is (up to 10 seconds)<\/li>\n<li id>What resolution (quality and dimensions)<\/li>\n<li id>Aspect ratio (wide, tall, square)<\/li>\n<\/ul>\n<h3 id>What You Can&#8217;t Make<\/h3>\n<p id>The API blocks certain things:<\/p>\n<ul id>\n<li id>Videos of real people, including celebrities<\/li>\n<li id>Characters from copyrighted movies or shows<\/li>\n<li id>Music that&#8217;s copyrighted<\/li>\n<li id>Anything not appropriate for teenagers<\/li>\n<li id>Images with people&#8217;s faces (for now)<\/li>\n<\/ul>\n<p id>If you try to generate these, the request fails with an error message.<\/p>\n<h3 id>Real Example<\/h3>\n<p id>Mattel, the toy company, uses the Sora 2 API. Their designers upload sketches of new toy ideas. <\/p>\n<p id>Sora 2 generates animated videos showing how the toy would look and move. <\/p>\n<p id>They can show these to decision-makers before building expensive physical prototypes.<\/p>\n<h3 id>Current Status<\/h3>\n<p id>The API is in preview. It works, but expect it to improve over time. New features will be added. Performance will get better.<\/p>\n<h3 id>Announcement #6: GPT-Realtime-Mini<\/h3>\n<p id>This is a new voice model. Smaller and cheaper than the existing realtime model.<\/p>\n<h3 id>The Key Number<\/h3>\n<p id>70% less expensive than the large realtime model. Same quality. Just cheaper.<\/p>\n<h3 id>What You Get<\/h3>\n<p id>All the same capabilities:<\/p>\n<ul id>\n<li id>Speech to speech (no text in the middle)<\/li>\n<li id>Natural conversation with appropriate emotion<\/li>\n<li id>Handles people interrupting<\/li>\n<li id>Multiple languages<\/li>\n<li id>Low delay between speaking and response<\/li>\n<\/ul>\n<h3 id>Use Cases<\/h3>\n<p id>This makes voice features affordable for:<\/p>\n<ul id>\n<li id>Voice assistants<\/li>\n<li id>Customer service bots<\/li>\n<li id>Accessibility tools for people with disabilities<\/li>\n<li id>Educational apps<\/li>\n<li id>Any application where talking to AI makes sense<\/li>\n<\/ul>\n<h3 id>When to Use Which<\/h3>\n<p id><strong id>Use Mini for:<\/strong> Most voice applications, especially if you&#8217;re doing it at scale and cost matters<\/p>\n<p id><strong id>Use the large realtime model for:<\/strong> When you need absolutely the best quality possible and cost isn&#8217;t the main concern<\/p>\n<h3 id>Availability<\/h3>\n<p id>Live in the API right now. Start using it today.<\/p>\n<h3 id>What All This Means Together<\/h3>\n<p id>Look at how these pieces connect:<\/p>\n<p id>Apps SDK lets you reach 800 million users. <\/p>\n<p id>AgentKit lets you build complex agents to power those apps. <\/p>\n<p id>Codex helps you write the code faster. <\/p>\n<p id>GPT-5 Pro gives you intelligence for hard problems. <\/p>\n<p id>Sora 2 lets you generate video content. <\/p>\n<p id>Realtime Mini makes voice affordable.<\/p>\n<p id>Everything works together to let you build and ship sophisticated software faster than before.<\/p>\n<h3 id>What&#8217;s Available Right Now<\/h3>\n<p id>You can start using today:<\/p>\n<ul id>\n<li id>Apps SDK (in preview mode for testing)<\/li>\n<li id>All four parts of AgentKit<\/li>\n<li id>Codex with all the new features<\/li>\n<li id>GPT-5 Pro through the API<\/li>\n<li id>Sora 2 API (in preview)<\/li>\n<li id>GPT-Realtime-Mini<\/li>\n<\/ul>\n<h3 id>What&#8217;s Coming Later This Year<\/h3>\n<p id>Expected in the remaining months of 2025:<\/p>\n<ul id>\n<li id>Apps SDK opens for all developer submissions<\/li>\n<li id>App directory where users can browse and search<\/li>\n<li id>Monetization options for apps including the commerce protocol<\/li>\n<li id>More improvements to Codex<\/li>\n<li id>Sora 2 API moving out of preview<\/li>\n<\/ul>\n<h3 id>What You Should Do Next<\/h3>\n<p id><strong id>If you&#8217;re a developer:<\/strong><\/p>\n<p id>Pick one thing from this list. <\/p>\n<p id>Try it this week. Build something small but real. <\/p>\n<p id>Learn how it actually works instead of just reading about it.<\/p>\n<p id>Agent Builder is probably the easiest to start with. <\/p>\n<p id>You can build something in an hour without writing code.<\/p>\n<p id><strong id>If you run a company:<\/strong><\/p>\n<p id>Think about one specific problem your business has. Could an agent solve it? <\/p>\n<p id>Could an app in ChatGPT reach new customers?<\/p>\n<p id>Run a small pilot. Measure the results. Expand what works.<\/p>\n<p id><strong id>If you&#8217;re just interested:<\/strong><\/p>\n<p id>Watch for apps appearing in ChatGPT. <\/p>\n<p id>Try them when they match what you&#8217;re doing. <\/p>\n<p id>See how different this feels from traditional software.<\/p>\n<p id>The shift is happening. The tools are real. The question is what gets built with them.<\/p>\n<div class=\"gop-cta\" style=\"margin:32px 0;padding:24px;border-radius:12px;background:#f5f5f5;text-align:center;\"><a href=\"https:\/\/godofprompt.ai\/prompt-library\" target=\"_blank\" rel=\"noopener\" style=\"display:inline-block;padding:14px 28px;background:#000;color:#fff;text-decoration:none;border-radius:8px;font-weight:600;\">Discover The Biggest AI Prompt Library By God Of Prompt<\/a><\/div>\n","protected":false},"excerpt":{"rendered":"<p>Everything announced at OpenAI DevDay 2025: build apps for 800M users, deploy AI agents with AgentKit, use Codex for coding, access new AI models in API.<\/p>\n","protected":false},"author":1,"featured_media":4641,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[3],"tags":[21],"class_list":["post-4642","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ai-tools","tag-tag-chatgpt"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v27.5 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>OpenAI DevDay 2025: Everything Announced and What It All Means | God of Prompt<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/godofprompt.ai\/blog\/openai-devday-2025\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"OpenAI DevDay 2025: Everything Announced and What It All Means | God of Prompt\" \/>\n<meta property=\"og:description\" content=\"Everything announced at OpenAI DevDay 2025: build apps for 800M users, deploy AI agents with AgentKit, use Codex for coding, access new AI models in API.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/godofprompt.ai\/blog\/openai-devday-2025\/\" \/>\n<meta property=\"og:site_name\" content=\"God of Prompt\" \/>\n<meta property=\"article:published_time\" content=\"2025-10-06T00:00:00+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/godofprompt.ai\/blog\/wp-content\/uploads\/2026\/05\/og-image-1200x630-1.jpg\" \/>\n\t<meta property=\"og:image:width\" content=\"1200\" \/>\n\t<meta property=\"og:image:height\" content=\"630\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/jpeg\" \/>\n<meta name=\"author\" content=\"Robert Youssef\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:creator\" content=\"@https:\/\/x.com\/rryssf\" \/>\n<meta name=\"twitter:site\" content=\"@godofprompt\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Robert Youssef\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"19 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/openai-devday-2025\\\/#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/openai-devday-2025\\\/\"},\"author\":{\"name\":\"Robert Youssef\",\"@id\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/#\\\/schema\\\/person\\\/d50f21f5201cf68185421f5fd87ed94f\"},\"headline\":\"OpenAI DevDay 2025: Everything Announced and What It All Means\",\"datePublished\":\"2025-10-06T00:00:00+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/openai-devday-2025\\\/\"},\"wordCount\":3918,\"publisher\":{\"@id\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/#organization\"},\"image\":{\"@id\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/openai-devday-2025\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/wp-content\\\/uploads\\\/2026\\\/05\\\/69ea6cba6c0e633fc8d26e61_68e44d658c1b6abb9d39313c_OpenAI-DevDay-2025-Complete-Summary-of-Every-2.avif\",\"keywords\":[\"ChatGPT\"],\"articleSection\":[\"AI Tool Tutorials\"],\"inLanguage\":\"en-US\"},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/openai-devday-2025\\\/\",\"url\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/openai-devday-2025\\\/\",\"name\":\"OpenAI DevDay 2025: Everything Announced and What It All Means | God of Prompt\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/openai-devday-2025\\\/#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/openai-devday-2025\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/wp-content\\\/uploads\\\/2026\\\/05\\\/69ea6cba6c0e633fc8d26e61_68e44d658c1b6abb9d39313c_OpenAI-DevDay-2025-Complete-Summary-of-Every-2.avif\",\"datePublished\":\"2025-10-06T00:00:00+00:00\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/openai-devday-2025\\\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/openai-devday-2025\\\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/openai-devday-2025\\\/#primaryimage\",\"url\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/wp-content\\\/uploads\\\/2026\\\/05\\\/69ea6cba6c0e633fc8d26e61_68e44d658c1b6abb9d39313c_OpenAI-DevDay-2025-Complete-Summary-of-Every-2.avif\",\"contentUrl\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/wp-content\\\/uploads\\\/2026\\\/05\\\/69ea6cba6c0e633fc8d26e61_68e44d658c1b6abb9d39313c_OpenAI-DevDay-2025-Complete-Summary-of-Every-2.avif\",\"width\":550,\"height\":308,\"caption\":\"OpenAI DevDay 2025: Everything Announced and What It All Means\"},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/openai-devday-2025\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"OpenAI DevDay 2025: Everything Announced and What It All Means\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/#website\",\"url\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/\",\"name\":\"God of Prompt\",\"description\":\"AI prompts, guides &amp; playbooks for ChatGPT, Claude, Gemini &amp; Midjourney\",\"publisher\":{\"@id\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/#organization\",\"name\":\"God of Prompt\",\"url\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/#\\\/schema\\\/logo\\\/image\\\/\",\"url\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/wp-content\\\/uploads\\\/2026\\\/05\\\/gop-logo.png\",\"contentUrl\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/wp-content\\\/uploads\\\/2026\\\/05\\\/gop-logo.png\",\"width\":512,\"height\":512,\"caption\":\"God of Prompt\"},\"image\":{\"@id\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/#\\\/schema\\\/logo\\\/image\\\/\"},\"sameAs\":[\"https:\\\/\\\/x.com\\\/godofprompt\",\"https:\\\/\\\/www.linkedin.com\\\/company\\\/god-of-prompt\\\/\",\"https:\\\/\\\/www.youtube.com\\\/@god-of-prompt\",\"https:\\\/\\\/www.instagram.com\\\/godofprompt\\\/\"],\"description\":\"God of Prompt is the AI prompt platform trusted by 100,000+ marketers, founders, and creators. We publish prompts, guides, and playbooks for ChatGPT, Claude, Gemini, and Midjourney.\"},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/#\\\/schema\\\/person\\\/d50f21f5201cf68185421f5fd87ed94f\",\"name\":\"Robert Youssef\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/d48b5a1e20bcb1d5a09591608fd744bc4303937062c5cbd00961fe65302db773?s=96&d=mm&r=g\",\"url\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/d48b5a1e20bcb1d5a09591608fd744bc4303937062c5cbd00961fe65302db773?s=96&d=mm&r=g\",\"contentUrl\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/d48b5a1e20bcb1d5a09591608fd744bc4303937062c5cbd00961fe65302db773?s=96&d=mm&r=g\",\"caption\":\"Robert Youssef\"},\"description\":\"The Missing Link I come from architecture and urban planning, designing systems that should have created leverage&mdash;transit networks, resource flows, development infrastructure. This work taught me how things should scale. When I shifted to helping businesses automate and implement AI, I kept seeing the same gap everywhere. Businesses had the technology. They had the need. But they were missing the layer in between&mdash;the infrastructure for how to actually communicate with AI. Developers spoke in functions. Clients spoke in outcomes. AI spoke in&hellip; whatever you prompted it to speak in. Nobody had a shared language. No protocols. No architecture. The Infrastructure Layer With generative AI becoming so essential, I stopped seeing AI as a tool and started seeing it as territory that needed architecture. People were treating it like a magic search bar. Ask once, get disappointed, move on. They were standing in front of a transit system but couldn&rsquo;t read the map. I realized: They don&rsquo;t need better AI. They need better infrastructure between them and AI. Prompts aren&rsquo;t requests&mdash;they&rsquo;re protocols. Communication architecture. The same thinking I used mapping resource flows in cities applied perfectly to designing how humans should interact with intelligence. Building the System @godofprompt became that infrastructure layer. Not a course. Not a tool. An intelligent system for how information should flow between human thinking and AI capability. Same principles that prevented scope creep in urban development now prevent prompt failures. Same patterns that identified bottlenecks in city budgets now identify bottlenecks in AI workflows. Turns out you don&rsquo;t need a bigger budget or better AI. You need someone who knows how to design the space between question and answer. That&rsquo;s AI architecture for me.\",\"sameAs\":[\"https:\\\/\\\/www.linkedin.com\\\/in\\\/rryssf\\\/\",\"https:\\\/\\\/x.com\\\/https:\\\/\\\/x.com\\\/rryssf\"],\"url\":\"https:\\\/\\\/godofprompt.ai\\\/blog\\\/author\\\/robert-youssef\\\/\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"OpenAI DevDay 2025: Everything Announced and What It All Means | God of Prompt","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/godofprompt.ai\/blog\/openai-devday-2025\/","og_locale":"en_US","og_type":"article","og_title":"OpenAI DevDay 2025: Everything Announced and What It All Means | God of Prompt","og_description":"Everything announced at OpenAI DevDay 2025: build apps for 800M users, deploy AI agents with AgentKit, use Codex for coding, access new AI models in API.","og_url":"https:\/\/godofprompt.ai\/blog\/openai-devday-2025\/","og_site_name":"God of Prompt","article_published_time":"2025-10-06T00:00:00+00:00","og_image":[{"width":1200,"height":630,"url":"https:\/\/godofprompt.ai\/blog\/wp-content\/uploads\/2026\/05\/og-image-1200x630-1.jpg","type":"image\/jpeg"}],"author":"Robert Youssef","twitter_card":"summary_large_image","twitter_creator":"@https:\/\/x.com\/rryssf","twitter_site":"@godofprompt","twitter_misc":{"Written by":"Robert Youssef","Est. reading time":"19 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/godofprompt.ai\/blog\/openai-devday-2025\/#article","isPartOf":{"@id":"https:\/\/godofprompt.ai\/blog\/openai-devday-2025\/"},"author":{"name":"Robert Youssef","@id":"https:\/\/godofprompt.ai\/blog\/#\/schema\/person\/d50f21f5201cf68185421f5fd87ed94f"},"headline":"OpenAI DevDay 2025: Everything Announced and What It All Means","datePublished":"2025-10-06T00:00:00+00:00","mainEntityOfPage":{"@id":"https:\/\/godofprompt.ai\/blog\/openai-devday-2025\/"},"wordCount":3918,"publisher":{"@id":"https:\/\/godofprompt.ai\/blog\/#organization"},"image":{"@id":"https:\/\/godofprompt.ai\/blog\/openai-devday-2025\/#primaryimage"},"thumbnailUrl":"https:\/\/godofprompt.ai\/blog\/wp-content\/uploads\/2026\/05\/69ea6cba6c0e633fc8d26e61_68e44d658c1b6abb9d39313c_OpenAI-DevDay-2025-Complete-Summary-of-Every-2.avif","keywords":["ChatGPT"],"articleSection":["AI Tool Tutorials"],"inLanguage":"en-US"},{"@type":"WebPage","@id":"https:\/\/godofprompt.ai\/blog\/openai-devday-2025\/","url":"https:\/\/godofprompt.ai\/blog\/openai-devday-2025\/","name":"OpenAI DevDay 2025: Everything Announced and What It All Means | God of Prompt","isPartOf":{"@id":"https:\/\/godofprompt.ai\/blog\/#website"},"primaryImageOfPage":{"@id":"https:\/\/godofprompt.ai\/blog\/openai-devday-2025\/#primaryimage"},"image":{"@id":"https:\/\/godofprompt.ai\/blog\/openai-devday-2025\/#primaryimage"},"thumbnailUrl":"https:\/\/godofprompt.ai\/blog\/wp-content\/uploads\/2026\/05\/69ea6cba6c0e633fc8d26e61_68e44d658c1b6abb9d39313c_OpenAI-DevDay-2025-Complete-Summary-of-Every-2.avif","datePublished":"2025-10-06T00:00:00+00:00","breadcrumb":{"@id":"https:\/\/godofprompt.ai\/blog\/openai-devday-2025\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/godofprompt.ai\/blog\/openai-devday-2025\/"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/godofprompt.ai\/blog\/openai-devday-2025\/#primaryimage","url":"https:\/\/godofprompt.ai\/blog\/wp-content\/uploads\/2026\/05\/69ea6cba6c0e633fc8d26e61_68e44d658c1b6abb9d39313c_OpenAI-DevDay-2025-Complete-Summary-of-Every-2.avif","contentUrl":"https:\/\/godofprompt.ai\/blog\/wp-content\/uploads\/2026\/05\/69ea6cba6c0e633fc8d26e61_68e44d658c1b6abb9d39313c_OpenAI-DevDay-2025-Complete-Summary-of-Every-2.avif","width":550,"height":308,"caption":"OpenAI DevDay 2025: Everything Announced and What It All Means"},{"@type":"BreadcrumbList","@id":"https:\/\/godofprompt.ai\/blog\/openai-devday-2025\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/godofprompt.ai\/blog\/"},{"@type":"ListItem","position":2,"name":"OpenAI DevDay 2025: Everything Announced and What It All Means"}]},{"@type":"WebSite","@id":"https:\/\/godofprompt.ai\/blog\/#website","url":"https:\/\/godofprompt.ai\/blog\/","name":"God of Prompt","description":"AI prompts, guides &amp; playbooks for ChatGPT, Claude, Gemini &amp; Midjourney","publisher":{"@id":"https:\/\/godofprompt.ai\/blog\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/godofprompt.ai\/blog\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Organization","@id":"https:\/\/godofprompt.ai\/blog\/#organization","name":"God of Prompt","url":"https:\/\/godofprompt.ai\/blog\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/godofprompt.ai\/blog\/#\/schema\/logo\/image\/","url":"https:\/\/godofprompt.ai\/blog\/wp-content\/uploads\/2026\/05\/gop-logo.png","contentUrl":"https:\/\/godofprompt.ai\/blog\/wp-content\/uploads\/2026\/05\/gop-logo.png","width":512,"height":512,"caption":"God of Prompt"},"image":{"@id":"https:\/\/godofprompt.ai\/blog\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/x.com\/godofprompt","https:\/\/www.linkedin.com\/company\/god-of-prompt\/","https:\/\/www.youtube.com\/@god-of-prompt","https:\/\/www.instagram.com\/godofprompt\/"],"description":"God of Prompt is the AI prompt platform trusted by 100,000+ marketers, founders, and creators. We publish prompts, guides, and playbooks for ChatGPT, Claude, Gemini, and Midjourney."},{"@type":"Person","@id":"https:\/\/godofprompt.ai\/blog\/#\/schema\/person\/d50f21f5201cf68185421f5fd87ed94f","name":"Robert Youssef","image":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/secure.gravatar.com\/avatar\/d48b5a1e20bcb1d5a09591608fd744bc4303937062c5cbd00961fe65302db773?s=96&d=mm&r=g","url":"https:\/\/secure.gravatar.com\/avatar\/d48b5a1e20bcb1d5a09591608fd744bc4303937062c5cbd00961fe65302db773?s=96&d=mm&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/d48b5a1e20bcb1d5a09591608fd744bc4303937062c5cbd00961fe65302db773?s=96&d=mm&r=g","caption":"Robert Youssef"},"description":"The Missing Link I come from architecture and urban planning, designing systems that should have created leverage&mdash;transit networks, resource flows, development infrastructure. This work taught me how things should scale. When I shifted to helping businesses automate and implement AI, I kept seeing the same gap everywhere. Businesses had the technology. They had the need. But they were missing the layer in between&mdash;the infrastructure for how to actually communicate with AI. Developers spoke in functions. Clients spoke in outcomes. AI spoke in&hellip; whatever you prompted it to speak in. Nobody had a shared language. No protocols. No architecture. The Infrastructure Layer With generative AI becoming so essential, I stopped seeing AI as a tool and started seeing it as territory that needed architecture. People were treating it like a magic search bar. Ask once, get disappointed, move on. They were standing in front of a transit system but couldn&rsquo;t read the map. I realized: They don&rsquo;t need better AI. They need better infrastructure between them and AI. Prompts aren&rsquo;t requests&mdash;they&rsquo;re protocols. Communication architecture. The same thinking I used mapping resource flows in cities applied perfectly to designing how humans should interact with intelligence. Building the System @godofprompt became that infrastructure layer. Not a course. Not a tool. An intelligent system for how information should flow between human thinking and AI capability. Same principles that prevented scope creep in urban development now prevent prompt failures. Same patterns that identified bottlenecks in city budgets now identify bottlenecks in AI workflows. Turns out you don&rsquo;t need a bigger budget or better AI. You need someone who knows how to design the space between question and answer. That&rsquo;s AI architecture for me.","sameAs":["https:\/\/www.linkedin.com\/in\/rryssf\/","https:\/\/x.com\/https:\/\/x.com\/rryssf"],"url":"https:\/\/godofprompt.ai\/blog\/author\/robert-youssef\/"}]}},"_links":{"self":[{"href":"https:\/\/godofprompt.ai\/blog\/wp-json\/wp\/v2\/posts\/4642","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/godofprompt.ai\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/godofprompt.ai\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/godofprompt.ai\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/godofprompt.ai\/blog\/wp-json\/wp\/v2\/comments?post=4642"}],"version-history":[{"count":0,"href":"https:\/\/godofprompt.ai\/blog\/wp-json\/wp\/v2\/posts\/4642\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/godofprompt.ai\/blog\/wp-json\/wp\/v2\/media\/4641"}],"wp:attachment":[{"href":"https:\/\/godofprompt.ai\/blog\/wp-json\/wp\/v2\/media?parent=4642"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/godofprompt.ai\/blog\/wp-json\/wp\/v2\/categories?post=4642"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/godofprompt.ai\/blog\/wp-json\/wp\/v2\/tags?post=4642"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}