Inspired by this story? Turn the idea into your own AI art in seconds — free to start, no card required.
Start creating free
Anthropic has upgraded Claude's voice mode to run on its Opus and Sonnet models — not just the faster but less powerful Haiku — and added the ability to take real actions inside Gmail, Slack, and Canva.
When Anthropic launched voice mode last year, the focus was on fluid, natural conversation — and Haiku's speed made it a reasonable fit for that use case. The tradeoff was capability: Haiku handles fast back-and-forth well, but it isn't the model you'd reach for when accuracy, nuance, or complex reasoning matters.
Shifting to Opus and Sonnet changes that calculus. According to The Verge, voice mode is now available across all three model tiers, with Opus and Sonnet unlocking meaningfully stronger reasoning behind the spoken interface. That matters most when a voice command isn't just a question but a task — rescheduling a calendar event, drafting a reply in Gmail, or updating a Canva design.
Claude voice mode now supports Opus and Sonnet models, enabling agentic actions in third-party apps.
Image: The Verge / The Verge AI
For creators who use Claude as part of a broader workflow — briefing a project in Slack, iterating on copy inside Canva, managing correspondence in Gmail — the practical shift is significant. You can now speak a multi-step instruction and have Claude execute it inside the app, rather than copy-pasting between a chat window and whatever tool you're actually working in.
The model upgrade isn't cosmetic. Haiku is optimized for latency; Opus is Anthropic's flagship reasoning model. Running voice mode on Opus means the same interface that felt useful for quick Q&A can now handle tasks that require judgment — parsing an ambiguous email thread, deciding what to prioritize in a Slack channel, or generating a Canva layout from a spoken description.
According to TechCrunch, the new voice model will let you reschedule meetings or draft emails directly through voice commands — a concrete demonstration of the agentic direction Anthropic is pushing. The distinction between "talking to an AI" and "having an AI work in your tools" is collapsing.
Claude's upgraded voice mode spans Haiku, Sonnet, and Opus, with the higher tiers enabling more complex task execution.
Image: The Verge / The Verge AI
For AI-art creators, Canva is the most immediately relevant integration. Canva has been steadily absorbing AI image tools — its own text-to-image features, background removal, and generative fill — and Claude voice mode now sits on top of that stack. Describing a layout change or asking Claude to generate and place an image element by voice, without switching to a separate chat interface, reduces friction in a workflow that already involves a lot of context-switching.
This upgrade fits a broader pattern across the AI industry: voice interfaces are moving from retrieval ("what is X?") to execution ("do X"). The Canva integration is a direct example of how that plays out for visual creators — a spoken prompt can now trigger a generative action inside the design tool itself, not just return a text description you then have to act on manually.
Anthropists building characters or creative projects in platforms like Charmloop's AI generator will recognize the workflow pressure this addresses: the more capable the underlying model, the more useful the shortcut. Running voice commands on Opus rather than Haiku means the spoken interface can actually handle the kind of nuanced, multi-step creative direction that visual work demands.
Anthropic hasn't announced pricing changes tied to the model upgrade, so whether Opus voice access requires a higher subscription tier remains worth checking before building it into a workflow.
