Skip to main content
silk is rumik ai’s speech api. it turns text into natural, expressive speech over a simple http call or a real-time websocket stream — in the original wav and raw pcm defaults, or as opus, pcm, mulaw, alaw, or mp3 — and runs conversational voice agents that listen and reply, in the browser or on the phone, which you can build and manage from code. pick the model that fits your use case:

silk muga 1

our more expressive model. steer delivery with a tone tag like [happy] and inline events like <laugh>.

silk mulberry 1.5

our faster model. steer with a natural-language description, or pick a preset studio voice.

silk mulberry 1.6

our multilingual model. speaks 22 indian languages in their own scripts, steered by style, accent and pace.

​
start here

quickstart

get a key and synthesize your first clip in three steps.

prompting guide

steer muga and mulberry with tones, tags, and descriptions.

stream in real time

low-latency playback over websocket.

choose an audio format

receive opus, pcm, mulaw, alaw, or mp3 from every delivery mode.

run a voice agent

start a conversational agent from your app — in the browser, over a socket, or on the phone.

build agents with the api

create agents, pick voices, rent numbers and place calls — everything the playground does, from code.

build a voice agent

drop rumik into a pipecat pipeline with pipecat-rumik.

api reference

endpoints, request fields, and a live playground.

​
your account

get an api key

create and manage keys in your dashboard.

open dashboard

view usage, current spend, and your plan details.

​
audio format

omit audio_format for the original behavior: http returns a 24 khz mono wav, and websocket sessions stream raw 24 khz mono signed 16-bit pcm. set audio_format to opus, pcm, mulaw, alaw, or mp3 to receive that format instead. see audio formats for exact content types and examples.

​
status

live api uptime and incident history are at silk-api.statuspage.io.