npx @buildinternet/releases get openaiRecently shipped Ultrafast mode for GPT-6.1 Sol across the API, Codex, and ChatGPT Work, and rolled out GPT-6 with Intelligent UI in ChatGPT.
Ultrafast mode now spans GPT-6.1 Sol.
service_tier: "ultrafast" shortens the gap between generated output tokens, with global processing and US and EU residency.GPT-6 and Intelligent UI reached ChatGPT globally, with faster responses and interactive visuals you can explore and use directly.
Decisions API turns text and images into typed answers. In beta with gpt-6-luna, it returns structured output roughly 10x faster than the Responses API, and SDKs added standalone support.
Agents API added computer use, letting agents complete tasks in an OpenAI-hosted browser with website approvals and sign-in handled by your application. Dots launched as a proactive assistant that keeps working across projects and everyday tasks.
Access and compliance.
Codex CLI iterated on session control. Opt-in instant_interrupt lets new input steer Codex mid-response, /mcp login signs into MCP servers from the terminal, and voice conversations gained device and channel selection.
Also shipped: GPT-6.1 Sol at one-fifth of Astra's standard token prices with Multi-agent beta, now the default in Codex and Bedrock catalogs; Python's browser and computer toolsets; College Planner, flashcards, and quizzes for ChatGPT for Teens; a visual ad format in ChatGPT.
Image requests and responses now include safety identifiers, and the Decisions cache breakpoint parameter type is shared across the API.
Image API requests and responses now include safety identifiers.
Adds prompt-based voice creation, regional processing residency options, response error diagnostics and model notices, and safety identifiers on image requests and responses, plus project API key scope updates. Also fixes typed-header request option isolation and upload handling.
The API now supports prompt-based voice creation, cache settings and key options, regional processing residency options, and response error diagnostics with model notices. Live sessions gain audio fixture playback and Decisions support, and Realtime response error messages are now declared.
Adds prompt-based voice creation and Decisions prompt caching options, along with Decisions support and audio fixture playback in Live sessions, project API key scope updates, and response error diagnostics and model notices. Fixed Realtime response error message declarations.
ChatGPT Slides lets users create presentations from a conversation, files, or an outline and refine text, images, shapes, and charts directly or by asking ChatGPT. Teammates can edit the same deck in real time and leave shared comments, with present-from-ChatGPT and export to PowerPoint, PDF, or Google Slides on web and PowerPoint or Keynote in the desktop app.
GPT-6 Pro in the chat tab now runs multiple agents that can work on different parts of a task in parallel, with progress visible in the conversation and completed tasks openable to see what each agent did. Responses are faster and plugin and tool use is improved, rolling out to eligible Pro users.
Sites admin preview is rolling out to eligible Enterprise workspaces so admins can review who can create and share Sites before enabling them for members, with public publishing still off by default. New Sites in US or EU data residency workspaces automatically use the workspace's existing region for supported storage and hosting, and tunnels are rolling out to connect Sites to private-network services without making them public.
GPT-Live (gpt-live-1 and gpt-6-luna) now has native Decisions integration, letting apps act while the user is speaking or react to Live output. Live can invoke Decisions through native delegation, or decisions can be triggered from input/output transcript deltas.
GPT-Live-1.1 and the smaller GPT-Live-1.1 Quip are rolling out in ChatGPT Voice with more natural accents across languages and a wider range of short phrases indicating when Live is searching. New Live voices cover French, German, Hindi, Japanese, Korean, and Mexican Spanish, plus Bubbie in English, selectable under Settings > Voice.
There is no longer a fixed limit on the number of Sites you can create, though usage and storage limits continue to apply.
Applications are now open to life sciences labs at U.S. universities, with approved labs receiving up to 30 ChatGPT Business Standard seats at no cost for up to 12 months and GPT-Rosalind access subject to trusted-access approval. Principal investigators or lab leads apply on behalf of their lab, and labs covered by an OpenAI Enterprise agreement should use their university's access process.
Workspace admins can make plugins available for installation, install for the workspace, or disable them in bulk, up to 25 at a time, in the Admin console. Role-based customization is available when RBAC is enabled, for Enterprise, Healthcare, and Edu workspaces.
New /config panel searches preferences across categories and flags which changes need a restart, and sending a follow-up while Codex is responding now steers it immediately, with interrupts stopping active Code Mode cells and nested tool calls. Also restarts managed background servers from /daemon, adds a configurable leader key in /keymap, and enables Daybreak by default for eligible accounts.
Workspace administrators can now select up to 25 plugins in the Admin console to make available for installation, install for the workspace, or disable, with dependency review before confirming. Available in Plugins by selecting the plugins and choosing Edit.