Sunesis for developers

Built by developers, for developers.

A clean REST + streaming API, typed SDKs in three languages, webhooks that just fire, and docs that answer the question before you have to ask it.

API request 200 OK

POST /v1/voice/stream

{
  "voice": "aria",
  "text": "Hello, so glad you're here.",
  "emotion": "warm"
}

Response · audio/pcm

chunk 1 · 4,096 bytes · streaming…

Auth

Bearer

Format

PCM

SDK

TypeScript

180 ms to first byte

Everything is one call away.

VoiceVideoDubbingAgentsWebhooks
What you get

A small API surface, on purpose.

Not a pile of AI endpoints with a docs problem. A focused surface for voice, character, and agent experiences, designed to be understood in an afternoon.

One key to voice, video, and agents

Bearer auth and a streaming endpoint return expressive audio while the words are still being spoken. Render talking characters from a script, open live agent sessions with a single call, and run dubbing on the same pipeline — the surface area stays small enough to hold in your head.

Bearer authExpressive streamingRealtime sessions
By the numbers
<60s

from signup to your first streamed audio, SDK or raw REST

Webhooks that just fire

Signed, retried, and idempotent. Know the moment a render, clone, or dub finishes — without polling, with every signature verifiable.

Typed SDKs, maintained by us

Idiomatic TypeScript, Python, and Go, versioned in lockstep with the API. Raw REST works everywhere the SDKs don’t.

Built to stay up

Multi-region, edge-terminated, 99.98% uptime, with a public status page and incident webhooks so you hear about problems from us first.

A focused developer at a minimal desk with a laptop showing soft code

Streaming voice in a lunch break, in production by Friday.

Platform engineer

Ship day

Signup to production before standup.

One key, a streaming endpoint, typed SDKs, and webhooks that just fire. The surface area is small on purpose, so the first thing you build is the thing you ship.

<60s
To first streamed audio
3
Typed SDKs, in lockstep
99.98%
Multi-region uptime
From zero to first generation

Streaming in under a minute.

Auth with a single key. Stream over HTTP. Webhooks for long-running jobs. The quickstart gets you from signup to first audio before your coffee cools.

Read the quickstart
01

Get your API key

Sign up, grab a key from the dashboard, and authenticate with a single Bearer header.

02

Make your first call

Follow the quickstart to stream your first audio in under a minute — SDK or raw REST.

03

Ship to production

Scope keys, wire up webhooks for long-running jobs, and scale on the free tier's 10,000 monthly credits.

Docs that answer first

Every endpoint, parameter, and error code, with copy-paste examples in three languages.

Free tier, real limits

10,000 credits a month, no credit card. Generous enough to build and test for real.

Support that ships code

Real engineers on the other end — people who care about the quality of what you launch.

Sunesis Labs | AI Voice, Character & Conversation Agents