Misaidizi wa sauti kwa muda mrefu wamekuwa wakikabiliwa na kizuizi cha kuchosha. Ungeweza kuwauliza kuhusu hali ya hewa, kuweka saa, au kuandaa orodha ya nyimbo. Mara tu mazungumzo yanapoelekea kwenye jambo tata—kama vile kusahihisha makosa ya kodi, kupanga muundo wa dili la kibiashara, au kujaribu uimara wa mkakati wa bidhaa—mawasiliano hayo hushindwa. Msaidizi anaweza kukusikia kwa usahihi, lakini anakosa kina cha uwezo wa kufikiri ili kutatua tatizo pamoja nawe.
Anthropic inajaribu kurekebisha hilo. Kampuni hiyo imepanua hali ya sauti (voice mode) katika Claude kutoka model yake ya awali ya Haiku hadi mifumo yake yenye uwezo mkubwa zaidi, Opus na Sonnet. Hatua hii inaashiria jambo
Anthropic has also designed the experience to break down barriers between different modes of interaction. Users can now jump between text and voice within the same conversation without losing context. You might start typing a technical query at your desk, switch to voice while walking to a meeting, then return to text to paste a code snippet.
The system also allows switching between model tiers midstream. If you begin a casual brainstorming session with Haiku—taking advantage of its snappy responses—you can escalate the conversation to Opus when the discussion turns to rigorous technical execution. The context transfers over. The AI remembers what you said three turns ago, regardless of whether you typed it or spoke it, or whether you were talking to the fast model or the deep one.
This fluidity matters because real work is rarely linear. Some problems need speed; others need depth. Building a single interface where users can move between those gears saves cognitive overhead. It prevents the friction of opening new tabs, copying prompts, or re-explaining context.
Speaking the World's Languages
The expansion is not limited to English speakers. Anthropic has ended the English-only beta, adding official support for nine additional languages:
- European languages: French, German, Italian, Spanish, and Portuguese.
- Asian languages: Hindi, Indonesian, Japanese, and Korean.
This is not merely a localization checkbox. High-reasoning voice assistance in Korean or Hindi changes who can use these tools for professional work. A startup founder in Jakarta can voice-draft an investor update in Indonesian. A manufacturing analyst in Turin can interrogate supply-chain data in Italian. The cognitive power of Opus becomes accessible to anyone who thinks more naturally in their native tongue than in English.
In the broader market for AI assistants, this multilingual rollout—paired with the reasoning capabilities of the top-tier models—sharpens Claude's competitive edge. Most voice assistants have historically treated non-English markets as afterthoughts, layering translation onto shallow reasoning. Anthropic seems to be betting that global users want the same depth of thought in their own languages that English speakers have come to expect.