Anthropic is giving Claude a voice to take on ChatGPT

If you use Claude for deep work and coding, you probably know the frustration of having to stop walking just to peck out a long prompt on your phone screen.

While OpenAI users have spent months casually chatting with Advanced Voice Mode on their commutes, Claude has remained stubbornly text-only. That is finally about to change.

Anthropic is preparing to roll out a new voice mode inside its Claude chatbot that could launch as soon as this month, according to a report from Bloomberg.

Three voices hidden in the code

The move was first spotted by app researcher M1Astra, who uncovered references to an upcoming voice feature buried inside Claude's iOS app. Bloomberg subsequently confirmed the findings.

Anthropic is reportedly preparing three distinct English voices at launch, dubbed Airy, Mellow, and Buttery.

This isn't an overnight pivot. Last month, Anthropic Chief Product Officer Mike Krieger teased the capability in an interview with the Financial Times. He noted that the company was already testing voice prototypes internally, calling it a clearly useful modality.

The voice leak follows a busy streak for the AI lab. Anthropic recently unveiled a steep $200 per month plan aimed at power users, alongside a new research-focused tool launched earlier this week.

Why it matters

For Claude fans, this is the missing puzzle piece. Having Claude's superior reasoning and nuanced writing available hands-free turns the assistant into a real companion for brainstorming, interview prep, or untangling complex thoughts while on the go. That said, non-English speakers will likely have to wait, given the initial launch focuses strictly on three English personas.

For the AI landscape, OpenAI's consumer moat just got a little narrower. ChatGPT's Advanced Voice Mode has been one of its biggest practical differentiators outside of pure benchmark scores. If Anthropic can match that conversational fluidity while leveraging Claude's reputation for fewer hallucinations and better conversational tone, it gives power users far less reason to keep paying for both services.

That said, execution will be everything. Real-time voice is brutally expensive to run and notoriously unforgiving when it comes to latency and conversational interruptions. It remains to be seen whether Anthropic treats voice mode as a broad consumer hook or locks it behind higher subscription tiers to offset compute costs.