Pindo

Placeholder copy

Four primitives,
one API key.

Mock content while the developer story is being written. Everything on this page is a stand-in for the real thing.

  • Speech to Text

    POST /v1/transcribe

    Stream or upload audio and get back a transcript with word timings and speaker labels.

    Kinyarwanda · Swahili · Luganda · English · French

  • Text to Speech

    POST /v1/speak

    Turn text into natural local-language audio, with a choice of voices and speaking rates.

    12 voices · WAV, MP3, µ-law

  • Voice Agents

    POST /v1/agents/calls

    Describe a task, attach your tools, and let Pindo run the whole conversation over a phone call.

    Inbound · Outbound · SIP trunking

  • Language Understanding

    POST /v1/understand

    Pull intents, amounts, names and dates out of a transcript in a single call.

    Entities · Intents · Sentiment

Placeholder copy

From zero to a transcript
in about a minute.

Get a key

Sign in, create a project, copy the test key. No sales call in the way.

Install the SDK

Official clients for TypeScript, Python and Go, plus plain HTTP for everything else.

Ship it

Test keys are free up to 500 minutes a month, then usage-based pricing.
import { Pindo } from "@pindo/sdk";

const pindo = new Pindo({ apiKey: process.env.PINDO_API_KEY });

const { text } = await pindo.transcribe({
  audio: fs.createReadStream("call.wav"),
  language: "rw",
});

console.log(text);

Built to be
boring in production.

Mock numbers. Region-pinned inference, idempotent requests, webhooks with signed payloads, and a status page you can subscribe to.

5
Languages in production
~320ms
Median first-token latency
99.95%
API uptime, trailing 90 days
12M+
Minutes processed