Anthropic Extends Claude Voice Mode to Opus and Sonnet
Anthropic expanded Claude's voice mode to its Opus and Sonnet model tiers on July 23, according to The Verge, moving spoken conversations beyond the faster Haiku model. The Verge also reports that voice sessions can use connected apps including Gmail, Slack, and Canva. Anthropic's help page still labels voice mode a beta feature, while separate testing indicates access may be rolling out in stages.
Anthropic expanded Claude's voice mode to its Opus and Sonnet model tiers on July 23, according to The Verge. Until this update, Claude's spoken interface primarily routed conversations through Haiku, the company's faster model tier.
The change matters because voice conversations can now use models intended for more demanding reasoning and knowledge-work tasks. The Verge also reports that voice mode is extending into connected apps such as Gmail, Slack, and Canva, allowing users to bring permitted information from those services into spoken interactions.
Availability is still settling
The public evidence points to a staged rollout rather than a perfectly synchronized launch. The Verge reports that Opus and Sonnet are now available in voice mode. TestingCatalog separately observed working Opus and Sonnet routing behind a feature flag earlier on July 23, but said the feature was not yet public at the time of its test. Those accounts support the same underlying product change while differing on how broadly it had reached users.
Anthropic's own voice-mode documentation confirms that the beta feature is available across Claude plans on the web, iOS, and Android. It describes hands-free and push-to-talk modes, web search during voice conversations, and saved text transcripts, but the retrieved page does not yet specify which model tiers are available. That documentation gap is why the exact rollout stage should be treated cautiously.
What changes for users
Moving from Haiku-only routing to selectable Opus or Sonnet gives users a way to trade speed and usage consumption for stronger reasoning during a spoken session. It does not mean every account will see the selector immediately, and Anthropic has not published model-specific voice limits or performance measurements on the retrieved help page.
For teams evaluating voice assistants, the practical questions are now model availability by plan, latency under the larger models, connected-app permissions, and how quickly usage limits are consumed. The product update expands Claude's conversational surface, but its value will depend on how consistently the staged rollout reaches accounts and how clearly Anthropic documents those operating limits.
Key Points
- 1The Verge reports that Claude voice mode now supports Opus and Sonnet, expanding spoken conversations beyond Haiku-only routing.
- 2TestingCatalog observed the same model routing behind a feature flag earlier on July 23, indicating that account availability may be staged.
- 3Anthropic's help page confirms voice mode is a cross-plan beta but does not yet document model-specific access, limits, or performance.
Scoring Rationale
A useful product expansion that brings stronger Claude model tiers to voice interactions and connected-app workflows, tempered by staged availability and limited official model-specific documentation.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
