Skip to content
Smallest AI (Pulse)

Smallest AI Pulse speech-to-text for Bolna AI voice agents

Pulse is Smallest AI's speech-to-text model, and in a Bolna voice agent it streams the caller's speech to text with sub-100ms first-token latency so the agent can reply quickly. Pick it when latency is your main concern, or when you want a single Smallest stack for both listening and speaking.

How Smallest AI fits in the stack.

  • Best for

    Turning the caller's speech into text in real time, in their language.

  • Use this layer when

    You want to tune accuracy, latency, languages or vocabulary for your callers.

  • Connects to

    The phone call upstream, and the LLM that reads the text downstream.

The transcriber is the
listening layer.

Smallest AI turns the caller's speech into text as they talk, so the LLM can understand it and reply.

  1. Telephony

    Phone network

    • Exotel
    • Plivo
    • Twilio
  2. STT · You are here

    Listener

    • Smallest AI (Pulse)
  3. LLM

    Reasoning

    • OpenAI
    • Anthropic
    • Google Gemini
  4. TTS

    Voice

    • ElevenLabs
    • Sarvam AI (TTS)
    • Cartesia
  5. Tools

    Actions

    • Cal.com
    • n8n
    • AiSensy

How do I set up Smallest AI Pulse with Bolna?

Which Smallest AI models and languages does Bolna support?

ModelLanguagesFeatures
pulse21 individual languages plus regional auto-detect in streaming mode, including English, Hindi, Tamil, Telugu, Kannada, Malayalam, major European languages and East Asian languagesStreaming transcription, code-switching, speaker labels, word and sentence timestamps, automatic punctuation, built-in noise handling, keyword weights

For the full configuration, see the Smallest AI setup guide in the Bolna docs.

The Bolna docs don't list every Pulse language by name. For the full list, see Smallest AI's Pulse documentation.

What can you build with Smallest AI and Bolna?

  • Low-latency call centres

    Large outbound or inbound operations where transcription delay directly affects how natural the conversation feels. Pulse stays around 300ms even at 100 concurrent requests.

  • Calls in English and South Indian languages

    Transcribe conversations that move between English and Tamil, Telugu, Kannada or Malayalam within the same call.

  • Clean, speaker-aware transcripts

    Pulse separates agent and caller speech and adds punctuation and timestamps, which makes post-call review and data extraction easier.

Other providers in this layer.

See all 10 speech-to-text providers
  • Soniox

    Speech-to-text

    Soniox stt-rt-v5 transcribes callers for your Bolna voice agent with one multilingual model that auto-detects and code-switches, including Hinglish.

  • AssemblyAI

    Speech-to-text

    AssemblyAI Universal-3.6 Pro transcribes callers for your Bolna voice agent in 32 languages, with code-switching and keyword plus context biasing.

  • Azure Speech-to-Text

    Speech-to-text

    Azure Speech-to-Text transcribes callers in real time for a Bolna voice agent, with speaker diarization, language identification and phrase lists of up to 2,000 terms.

Keep exploring the voice stack.

See where Smallest AI fits in your workflow.

Walk through provider choice, stack trade-offs and the exact workflow you want Bolna to automate.

Ready for enterprise.

Secure, compliant and dependable, from the first pilot to full scale.

Contact Sales
  • 99.9% uptime

    Built for your busiest hour. Even if the model fails, the call doesn’t

  • Support and SLAs

    A dedicated team, with service levels shaped around your business

  • Security & compliance

    Data residency, encryption, access controls, full audit trails and more

  • Deploy in your cloud

    Run Bolna in your own cloud or on-premise, your data remains yours

Certified by

  • ISO 27001
  • DPDP compliant

Frequently Asked Questions

Pulse is designed for sub-100ms first-token latency at single concurrency, and stays around 300ms at 100 concurrent requests.