OpenAI brings GPT-Live voice agents to ChatGPT desktop, Work, and Codex

OpenAI announced ChatGPT Voice, a GPT-Live-powered capability that brings spoken interaction and live coordination to the ChatGPT experience on desktop and mobile. The update adds voice input and output across ChatGPT.com and mobile apps and, according to release notes, brings voice into the Windows desktop app while unifying Chat, ChatGPT Work, and Codex in the desktop experience.

What OpenAI announced

OpenAI described ChatGPT Voice as powered by GPT-Live, enabling the app to speak, listen, and coordinate tasks simultaneously. The company says the capability is rolling out broadly to iOS, Android, and ChatGPT.com, and the Windows desktop release notes confirm that the desktop app now includes voice and a unified experience for Chat, Work, and Codex. For details from OpenAI, see OpenAI's GPT-Live announcement.

How ChatGPT Voice and GPT-Live work in-app

OpenAI positions GPT-Live as a real-time orchestration layer that can operate conversationally: it can emit speech, process microphone input, and coordinate multiple in-app agents at the same time. The launch message and release notes state that users can control their computer and direct multiple agents running in ChatGPT Work or Codex using voice, which is intended to streamline hands-free workflows and developer tasks in the app.

Rollout and platform coverage

OpenAI confirmed the rollout covers ChatGPT on the web (ChatGPT.com) and native iOS and Android apps. Release notes for the Windows desktop app specifically call out the addition of voice and a consolidated desktop interface that includes Chat, Work, and Codex features. The company framed the release as a broad rollout rather than a limited beta, though availability may vary by account, region, or platform at first.

Pricing, privacy, and governance implications

OpenAI's announcement focuses on capability and platform rollout; it does not lay out new pricing details in the materials cited. Organizations should verify pricing and billing terms in OpenAI's product pages and account settings before adopting voice-driven workflows.

Voice agents introduce operational questions for enterprises and developers. Key considerations include microphone and device permissions, data retention and logging of voice interactions, access controls for multi-agent coordination, and compliance with industry-specific regulations. The announcement does not replace the need for enterprises to review privacy policies and governance controls before enabling voice features for production use. Organizations evaluating integration of GPT-Live voice agents into workflows can work with Scalevise on AI architecture, automation, and implementation to define governance, security, and deployment strategies.

Frequently Asked Questions

What platforms support ChatGPT Voice today?

OpenAI states ChatGPT Voice is rolling out on ChatGPT.com and the iOS and Android apps, and release notes confirm the Windows desktop app now includes voice and a unified Chat, Work, and Codex experience.

Is ChatGPT Voice powered by GPT-Live?

Yes. OpenAI announced that ChatGPT Voice is powered by GPT-Live, which enables simultaneous speaking, listening, and agent coordination inside the app.

Can I use voice to control agents in ChatGPT Work and Codex?

According to OpenAI, users can direct multiple agents running in ChatGPT Work or Codex by voice, and the desktop experience is intended to support that multi-agent coordination.

Does the announcement include pricing or tiers for voice features?

OpenAI's announcement and the cited release notes emphasize capability and rollout. They do not provide detailed pricing or tier information for ChatGPT Voice; customers should check OpenAI's official product pages and account dashboards for pricing updates.

What should enterprises do before enabling voice features?

Enterprises should review device permission settings, logging and retention policies, access controls for agent coordination, and regulatory compliance requirements. Pilot testing in a controlled environment and defining governance policies are recommended steps.

Conclusion

OpenAI's ChatGPT Voice brings real-time spoken interaction to ChatGPT across web, mobile, and the Windows desktop while integrating Chat, Work, and Codex in the desktop app. The feature is powered by GPT-Live and is positioned to accelerate hands-free workflows and multi-agent coordination, but organizations should confirm availability, pricing, and governance controls before broad adoption. Scalevise can assist teams that want help evaluating integration paths, governance, and automation for GPT-Live voice agents.