ChatGPT adds voice-driven agentic features to its mobile app

🕒 Published on Zendoric: September 25, 2026 · 03:17
✨ AI-generated · how it's made
OpenAI announced on Wednesday that it is bringing voice-based agentic features to the ChatGPT mobile app, letting users trigger workflows with voice commands, such as drafting documents or summarizing emails.
OpenAI announced on Wednesday that it is bringing voice-based agentic features to the ChatGPT mobile app, allowing users to trigger workflows through voice commands, such as drafting documents or summarizing emails.
According to the article, Plus and Pro subscribers will be able to use the Work tab on their phones to create a document, draft an email or summarize Slack messages, for example. These paying users can also carry out other tasks such as building websites, creating presentations, using the cloud browser and accessing other areas of ChatGPT, including finance. Users on the Free and Go plans, by contrast, will have access to plugins and connected apps, but not to these more advanced features.
OpenAI noted that ChatGPT voice conversations will now have richer text output, and that users will be able to switch easily between text and voice. In addition, it will be possible to start a conversation on mobile and pick it up later on desktop, pointing to a more seamless experience across devices.
The update comes at a time when more and more users are turning to voice to instruct AI assistants on complex tasks. The article recalls that in July OpenAI launched its conversational model GPT-Live, which it subsequently integrated into the desktop app to enable voice use in the Work tab or to build applications in the Codex tab. This mobile rollout extends those same capabilities to that environment.
The article also mentions, as competitive context, that Anthropic has recently made it easier to hand off tasks between mobile and desktop, in addition to merging its Cowork and Chat interfaces. OpenAI, by contrast, is for now keeping the chat and workspace sections of its product separate.
Taken together, the news reflects a broader trend toward AI assistants that operate agentically — that is, carrying out multi-step tasks on their own — and toward voice as an increasingly important interface for launching that kind of workflow, on both mobile and desktop.
🔗 Related on Zendoric


