Limit 1
The WaveNet adapter is non-streaming: Google returns the full clip before the first 200 ms PCM frame is yielded.
हिंदी · measured India voice evidence
Hindi AI voice: runtime path, real pack lines and measured latency
Hindi resolves to Bharat Essential, Google Cloud Text-to-Speech voice hi-IN-Wavenet-A, the hi-IN dedicated pack and the cascade engine. Three fresh production-adapter runs measured a 822 ms median to the first non-empty 8 kHz PCM audio chunk on 15 Aug 2026.
Measured 15 Aug 2026 · three runs · 8 kHz PCM16 · no provider fallback
Which pack, engine and vendor voice serve Hindi?
The runtime trace starts with BaseBusiness.get_voice_pack, continues through voice_config.resolve_pack and ends at voice_config.resolve_engine. The generic integration returned none from the generic business integration, resolve_pack selected the hi-IN dedicated pack, and resolve_engine selected cascade. The TTS factory then selected Google Cloud Text-to-Speech voice hi-IN-Wavenet-A, displayed as Kavya, while speech recognition remained Sarvam saaras:v3.
| Runtime check | Resolved value | Meaning |
|---|---|---|
| get_voice_pack | None from the generic business integration | A client-specific pack would override the generic registry result. |
| resolve_pack | hi-IN dedicated pack | Fixed lines and language rules come from a locale-specific pack. |
| voice_config | Bharat Essential · Google Cloud Text-to-Speech · hi-IN-Wavenet-A | The grade write-through pins the production TTS provider and voice identifier. |
| resolve_engine | cascade | The cascade engine uses separate STT, reasoning and TTS stages. |
| speech recognition | Sarvam saaras:v3 | The primary recognition locale is hi-IN. |
The runtime table names the actual Hindi call path rather than a language-support logo: generic integration result, resolved pack, sold grade, provider voice, engine and speech-recognition provider. A client integration can replace the generic pack, but the benchmark fixture used no client override.
What first-audio latency did Hindi measure?
Hindi measured 788, 822, 946 ms across three isolated production-adapter requests. The median was 822 ms. Each run synthesized “नमस्ते, मैं आपकी कैसे मदद कर सकती हूँ?” at 8 kHz PCM16, started the timer immediately before speak(text), stopped at the first non-empty audio chunk, verified the configured adapter had not failed over, and cancelled the iterator after that first chunk.
| Run | First audio | Audio format | Adapter result |
|---|---|---|---|
| Run 1 | 788 ms | 8 kHz PCM16 | Google Cloud Text-to-Speech; no fallback |
| Run 2 | 822 ms | 8 kHz PCM16 | Google Cloud Text-to-Speech; no fallback |
| Run 3 | 946 ms | 8 kHz PCM16 | Google Cloud Text-to-Speech; no fallback |
The latency table reports a median of 822 ms from three measurements on 15 Aug 2026. The number is a reproducible TTS-adapter read, not end-to-end call latency, a percentile, or an uptime promise. Caller speech, transcription, model generation, carrier delivery and handset playback remain outside the timer.
What register does the resolved Hindi pack use?
The dedicated Hindi pack uses polite service Hindi with माफ़ कीजिए and बताइएगा. The longer recovery line uses a feminine agent persona through ले लेती हूँ and भेज नहीं सकती.
| Pack event | Exact runtime string | Register read |
|---|---|---|
| Filler while a tool runs | एक पल | A compact neutral Hindi hold line. |
| Retry after unclear speech | माफ़ कीजिए, एक बार फिर बताइएगा? | Polite service Hindi rather than casual तू/तुम register. |
| Retract an unsupported action | माफ़ कीजिए, वो मैं यहाँ से भेज नहीं सकती। आपका नंबर ले लेती हूँ, कोई आपको भेज देगा। | Feminine agent form with a human fallback. |
The sample table copies three fixed strings from the hi-IN dedicated pack: the tool filler, unclear-speech retry and unsupported-action retraction. Hindi therefore has explicit runtime register evidence.
What are the honest Hindi limits?
Limit 2
The benchmark uses one short Hindi sentence and three isolated requests; 822 ms is a reproducible median, not a production percentile or SLA.
Limit 3
A dedicated Hindi pack does not prove every business name, English product name or regional accent will be pronounced correctly.
Limit 4
The first-audio benchmark measures TTS adapter latency only; it excludes caller speech recognition, model reasoning, carrier transit and handset playback.
Which evidence supports the Hindi hub?
The fresh GSC pull found 34 impressions for the existing Hindi voice-agent post, including 33 anonymised impressions, making Hindi the highest-demand language route. The existing language article remains the workflow and cost depth page; the multilingual category index provides cross-language buying context; and the product page describes the broader scheduled-language surface.
Other measured Indian-language hubs
The related hubs keep one route per language and never multiply languages by verticals. Every page publishes its own runtime trace, three dated latency reads and fixed-pack evidence.
Hindi voice questions
The answers preserve the runtime resolution, measured median, exact run values and pack limitation as standalone passages that do not depend on the tables.
Hindi resolves to Bharat Essential, the Google Cloud Text-to-Speech hi-IN-Wavenet-A voice, Sarvam saaras:v3 recognition and the cascade engine. The generic integration returned no client-specific pack, so resolve_pack selected the hi-IN dedicated pack on 15 Aug 2026.
Hindi measured 822 ms median first-audio latency across three fresh 8 kHz PCM16 adapter runs: 788, 822, 946 ms. The timer started immediately before the production adapter's speak call and stopped at the first non-empty audio chunk on 15 Aug 2026.
Hindi uses the hi-IN dedicated pack. The fixed pack strings shown on this page are copied from the runtime object rather than written as marketing examples.