Two big updates to the Responses API today.

🖇️ Connectors — Pull context from Gmail, Google Calendar, Dropbox, and more in a single API call.

💬 Conversations — Persist chat threads for your users, without running your own database.

More below: Build rich experiences with connectors to:

- Read emails
- Fetch calendar events
- Search files and chats

Gmail, Google Calendar, Drive, Dropbox, Teams, Outlook Calendar + Email, and SharePoint are available now.

They work with deep research, too!

platform.openai.com/docs/guides/to…

Six tips for coding with GPT-5:

Here’s a megathread of GPT-5 already at work:

https://x.com/cursor_ai/status/1953519582372831266

Two new additions to the API:

📚 Deep research
🪝 Webhooks o3-deep-research: platform.openai.com/docs/models/o3…
o4-mini-deep-research: platform.openai.com/docs/models/o4…

These models are the same post-trained o3 and o4-mini models that power deep research in ChatGPT.

They also support MCP (search/fetch) and Code Interpreter.

🇵 🇷 🇴 🇲 🇵 🇹 🇸 are an API primitive now!

Centrally manage, version, and optimize prompts. Use (and reuse!) prompts across the Playground, API, Evals, and Stored Completions. 🆕 The Prompt object is a new resource that can be referenced in the Responses API and our SDKs. Prompts can be preconfigured too—including tools, models, and messages. No more manual copying and pasting!

Evals now supports tool use. 🛠️

You can now use tools and Structured Outputs when completing eval runs, and evaluate tool calls based on the arguments passed and responses returned. This supports tools that are OpenAI-hosted, MCP, and non-hosted. Read more in our guides below. Web Search evaluation: cookbook.openai.com/examples/evalu…

We’re rolling out a few updates to Codex today:

1. Codex is rolling out to ChatGPT Plus users today. It includes generous usage limits for a limited time, but during periods of high demand, we might set rate limits for Plus users so that Codex remains widely available. 2. Next, our top requested feature: You can now give Codex access to the internet during task execution to install base dependencies, run tests that need external resources, upgrade or install packages needed to build new features, and more.

🆕 Four updates to building agents with OpenAI: Agents SDK in TypeScript, a new RealtimeAgent feature for voice agents, Traces support for the Realtime API, and improvements to our speech-to-speech model. The Agents SDK is now available in TypeScript and supports handoffs, guardrails, tracing, MCP, and other core agent primitives, just like the Python version.

openai.github.io/openai-agents-…

Image gen is now available in the API!

We’re launching gpt-image-1, making ChatGPT’s powerful image generation capabilities available to developers worldwide starting today.

✅ More accurate, high fidelity images
🎨 Diverse visual styles
✏️ Precise image editing
🌎 Rich world knowledge
🔤 Consistent text rendering

Check out the guide to start building: platform.openai.com/docs/guides/im… You can also use gpt-image-1 in the Playground to quickly iterate on prompts and images: platform.openai.com/playground/ima…

We're launching new tools to help developers build reliable and powerful AI agents. 🤖🔧

Timestamps:
01:54 Web search
02:41 File search
03:22 Computer use
04:07 Responses API
10:17 Agents SDK Our new API primitive: the Responses API. Combining the simplicity of Chat Completions with the tool-use of Assistants, this new foundation provides more flexibility in building agents. Web search, file search, or computer use are a couple lines of code!

platform.openai.com/docs/quickstar…

We’ve put together some best practices on using our o-series models:

🤔 Differences between our reasoning and non-reasoning models
💡 When to use reasoning models
💬 How to approach prompting reasoning models effectively GPT models focus on fast, cost‑efficient, straightforward task execution—“the workhorses.” 🐎

o‑series models excel at handling ambiguous, multi‑step tasks in domains such as math, engineering, legal, and finance—“the planners.” 🧠

We’ve put together a reference implementation for building and orchestrating agentic patterns using the Realtime API. You can use this repo to prototype a voice app using multi-agent flows in less than 20 minutes!

github.com/openai/openai-… Building with the Realtime API can be complex because of the low-latency, synchronous nature of voice interactions. This repo includes best practices we’ve learned for managing this complexity, like:

- Orchestrating agent handoffs (inspired by Swarm)
- Background escalation to o1 for advanced decision making
- Improving model instruction following by defining a state machine in the prompt
- Demos of applying these patterns to customer service and front desk use cases

We announced the Realtime API last week at DevDay SF. It's been amazing to see its adoption—here are some of the coolest examples we’ve seen so far. Let us know if we missed any in this thread!

https://x.com/jsngr/status/1843762165167325525

Today at DevDay SF, we’re launching a bunch of new capabilities to the OpenAI platform: 🗣️ Introducing the Realtime API—build speech-to-speech experiences into your applications. Like ChatGPT’s Advanced Voice, but for your own app. Rolling out in beta for developers on paid tiers. openai.com/index/introduc…

Introducing a series of updates to the Assistants API 🧵

With the new file search tool, you can quickly integrate knowledge retrieval, now allowing up to 10,000 files per assistant. It works with our new vector store objects for automated file parsing, chunking, and embedding. New token controls allow you to set maximum input and output tokens per run to manage costs. You can also choose how many recent messages to use for context truncation.

platform.openai.com/docs/assistant…

GPT-4 Turbo with Vision is now generally available in the API. Vision requests can now also use JSON mode and function calling.


Below are some great ways developers are building with vision. Drop yours in a reply 🧵platform.openai.com/docs/models/gp… Devin, built by @cognition_labs, is an AI software engineering assistant powered by GPT-4 Turbo that uses vision for a variety of coding tasks.