Skip to content
NewsProduct launch

ChatGPT Enterprise and Edu get file uploads and Projects context inside GPT-Live voice sessions

· by Pondero Newsdesk

The short version

OpenAI on August 7 extended GPT-Live for Enterprise, Edu, and Healthcare tiers to accept file uploads mid-call and pull in Projects context. GPT-Live also became the default voice experience for those workspaces, removing the Early Model Access prerequisite.

ChatGPT Enterprise and Edu get file uploads and Projects context inside GPT-Live voice sessions

A voice session in ChatGPT Enterprise could not access uploaded files or Projects context until August 7. OpenAI pushed both capabilities into GPT-Live that day for Enterprise, Edu, and Healthcare tiers, and simultaneously made GPT-Live the default voice experience for those workspaces, removing the Early Model Access setting that had previously been a prerequisite, per OpenAI's Enterprise and Edu release notes.

What changed

GPT-Live launched on July 8, 2026 as OpenAI's full-duplex voice architecture. Unlike the earlier Advanced Voice mode, it processes speech input and output at the same time, enabling natural interruptions and real-time back-and-forth without turn-based delays. On August 7, two capabilities that had been restricted to text sessions moved into that voice layer.

File upload support lets a user hand a document to ChatGPT mid-conversation and ask questions about its contents. Project integration draws on the active Project's history, uploaded sources, and custom instructions while a voice session runs. Both work through GPT-Live-1, the higher-capability version of the model set, per Releasebot's changelog aggregating OpenAI's release notes.

On the workspace side, Enterprise, Edu, and Healthcare owners who enable Voice now get GPT-Live without any additional configuration step. Early Model Access is no longer a prerequisite.

Why it matters

Before August 7, an analyst on a voice call could not pull in the document or Project instructions already loaded in their text session. Switching required stopping the voice conversation, opening a text window, and re-establishing context. Removing that step makes the voice interface practical for workflows where documents are the subject of the conversation, not background material.

Workspace administrators should note the default-voice change carries an operational consequence: any Enterprise or Edu tenant whose owner enables Voice will receive GPT-Live automatically. Organizations managing a phased voice rollout should audit their workspace Voice settings. Switching from Advanced Voice to GPT-Live is not a minor update; the underlying model architecture and interaction model are distinct.

Projects entering voice is notable in its own right. OpenAI's persistent workspace structure has been the company's main organizational layer since its late-2024 introduction. Extending Projects into voice signals that GPT-Live is being positioned as a full primary interface rather than a specialized supplementary mode.

What to watch next

OpenAI has not confirmed whether file uploads in voice or Projects-in-voice will extend to Plus, Pro, or free tiers. August 7's release notes apply specifically to Enterprise, Edu, and Healthcare workspaces. No comparative data on file-analysis accuracy or latency in voice versus text has been published. Independent testing of document handling inside GPT-Live will be the first meaningful signal for teams deciding whether to route document-heavy workflows to voice.

Sources