Claude's Voice Mode Grows Up: Opus and Sonnet Replace Haiku-Only
Anthropic upgraded Claude's voice mode beyond Haiku for the first time, adding Opus and Sonnet model choice, cross-app connectors, and support for ten more languages.
Anthropic has upgraded Claude's voice mode for the first time since launch, moving it off Haiku-only and onto a full model lineup — a fix for a real problem: voice mode previously routed every request through Claude's smallest model, which struggled on complex asks.
What changed
Paid users can now switch between Haiku, Sonnet, and Opus mid-conversation in voice mode. By default, voice mode picks up whichever model the user was last using in text chat, running its fastest available version — so the experience stays quick for simple exchanges but can hand off to a more capable model when a question demands it.
Connectors and language support
The update also adds:
- Cross-app automation with Gmail, Google Calendar, Slack, Canva, and Notion — letting voice mode actually take actions in connected tools, not just talk
- Support for ten major languages: English, French, German, Hindi, Indonesian, Italian, Japanese, Korean, Portuguese, and Spanish
Access tiers
- Free accounts: limited to a single connection, and every prompt runs through Haiku — though all supported languages remain available
- Paid plans: unlock Opus, Sonnet, and every connected tool
"Connectors" here means voice mode can trigger real actions in apps like Gmail or Slack mid-conversation — e.g. asking Claude out loud to check your calendar or draft an email — rather than only answering questions.
Why it matters
For everyday users: voice mode no longer feels artificially limited compared to text chat — the model doing the reasoning is now the same caliber you'd choose for a hard question in writing.
For productivity use cases: connectors turn voice mode from a Q&A tool into something closer to a spoken assistant that can actually act across your existing apps.
For the competitive landscape: this update lands the same month OpenAI shipped its own voice upgrades — the developer-facing gpt-realtime-2.1 API models and the consumer-facing GPT-Live — making voice quality and capability a genuine, active front between the two labs rather than a solved feature.
What to watch
- Whether free-tier users eventually get access to Sonnet or Opus in voice mode, even in limited form
- How many additional connectors Anthropic adds beyond the initial five apps
- Whether Claude's full-duplex or interruption handling narrows the gap with OpenAI's GPT-Live
Source: TechCrunch
NextGen AI Digest Editorial
Editorial Team
Reporting and analysis from the NextGen AI Digest newsroom — covering AI, agentic systems, SaaS, and the future of technology. Every piece is factual, sourced, and cited. Built and published by the team at Peaders.
Keep reading
The Fable 5 Reset: Why Anthropic's Most Powerful Model Went Dark and Came Back
Claude Fable 5 launched June 9, was forced offline by a U.S. export directive within three days, and returned with new cybersecurity classifiers. Inside the first government-ordered model recall.
The White House Is Now Shaping Who Gets Access to Frontier AI Models
A new federal program, Gold Eagle, puts the U.S. government in the loop on which partners get early access to frontier models from labs like OpenAI and Anthropic.
OpenAI Quietly Ships Faster Voice Models for Developers: GPT-Realtime-2.1
Ahead of the consumer-facing GPT-Live launch, OpenAI shipped gpt-realtime-2.1 and a mini variant to the API — cutting latency and adding reasoning effort controls for voice agents.