Bhāṣā

Last updated: May 2026

What we don't collect

Bhāṣā has no accounts, no login, and no server-side storage of your data. We do not store conversation transcripts, voice recordings, or usage data.

How your voice is processed

When you speak, your audio is transcribed in real time by Soniox. The text is sent to Anthropic (Claude) to generate a response. That response is converted to audio by Google Cloud Text-to-Speech. Each step is transient — nothing is retained once the request completes.

What stays on your device

Notes you save and your app settings — language, voice preference, and access code — are stored in your browser's local storage. This data never leaves your device and is not transmitted to us.

Third-party services

Your voice and text pass through three services:

• Soniox — real-time speech transcription
• Anthropic — AI response generation
• Google Cloud — text-to-speech audio

Each processes data transiently. We recommend reviewing their respective privacy policies for full detail.

Access code

Your access code authenticates API requests. It is not linked to your name, email, or any personal identifier.

Tracking

Bhāṣā uses no cookies, no analytics tools, and no third-party trackers.

Changes

We may update this policy as the product evolves. The date above reflects the most recent revision.

Contact

Questions? Reach us at hello@bhasha.app