Anthropic обновила голосовой режим Claude: перенос встреч и черновики писем

On July 23, 2026, Anthropic updated Claude's voice mode, connecting it to more capable models. Now voice can be used not just to ask questions, but to carry out work tasks — for example, rescheduling a meeting or drafting an email.

What voice Claude can now do

After the update from July 23, 2026, Claude's voice mode moves to more capable models and takes on concrete actions, not just conversation. As TechCrunch reports, the assistant helps reschedule a meeting and dictate an email draft — that is, it works with the user's calendar and mail directly through voice.

• Update date — July 23, 2026
• What changed — Claude's voice mode got more capable models
• Scenarios from the announcement — rescheduling a meeting and an email draft
• Focus — performing actions, not just answering questions

The key difference from previous versions is the shift from "talking to the assistant" to "instructing the assistant." The user describes a task verbally, and Claude carries it out in connected services, rather than simply reading an answer aloud. Anthropic positions the update as part of Claude's evolution toward practical work scenarios.

How this differs from voice search

Claude's voice mode in the version from July 23, 2026, is oriented toward actions rather than finding answers. Ordinary voice assistants answer a question and stop; the updated Claude, according to TechCrunch, carries a task through to a result — changes the time of a meeting or prepares the text of an email that only needs to be sent.

Voice interfaces at major AI assistants have in recent years evolved toward natural conversation; Claude's update shifts the emphasis to a completed action. This is the difference between an assistant that explains how to reschedule a meeting and an assistant that reschedules it.

This shift matters because routine operations — rescheduling a call, replying to an email, editing a schedule — take up a noticeable part of the workday. More capable models under the hood are needed precisely so the assistant more accurately understands the spoken formulation and doesn't make mistakes in details like the date, time, and recipient, where the cost of an error is high.

"Claude's new voice model will let you reschedule a meeting or draft an email," is how

TechCrunch describes the update.

Why this matters for Anthropic

For Anthropic, the voice mode update on July 23, 2026, is a step toward agentic scenarios, where Claude doesn't just answer but acts. The company has long promoted Claude as a work tool for business tasks, and voice with action execution fits into this line.

Anthropic has not yet disclosed which specific models are behind the update or on which plans the new voice mode is available — the announcement speaks of "more capable models" without specific names. This leaves open questions about speed, price, and language support, but sets the direction: voice is becoming a full-fledged interface to work services, rather than a separate feature for short queries.

The voice interface is becoming a field of competition among leading AI companies: a conversational mode exists at most major assistants, and the next frontier is not the quality of speech synthesis, but the ability to carry a spoken request through to a completed action. Claude's update from July 23, 2026, is Anthropic's bid for precisely this frontier.

What this means

The voice assistant is ceasing to be a demo and becoming a way to delegate routine work: the updated Claude performs actions in the calendar and mail on a spoken command. This is another signal that AI assistants in 2026 are moving from answers to actions — and voice is becoming one of the main interfaces for this transition.