Bowmark AIdocs

ElevenLabs

ElevenLabs' own documented REST API (api.elevenlabs.io) — converts text into spoken audio in an existing or newly-cloned voice, and clones a new voice…

ElevenLabs' own documented REST API (api.elevenlabs.io) — converts text into spoken audio in an existing or newly-cloned voice, and clones a new voice from caller-supplied audio samples (Instant Voice Cloning). Requires an ElevenLabs API key.

Domain: elevenlabs.io

Also known as: ElevenLabs, elevenlabs, elevenlabs.io, Eleven Labs, ElevenLabs Inc.

Prefer the capability

bowmark.text_to_speech covers this provider and routes around it when it is having a bad day. Reach for this page when you need elevenlabs.io specifically.

Call it directly

bowmark.providers.elevenlabs.synthesize(args: { text: string; voiceId?: string; modelId?: string }): Promise<elevenlabsSynthesisResult>
bowmark.providers.elevenlabs.cloneVoice(args: { name: string; sampleUrls: string[]; description?: string }): Promise<elevenlabsVoice>

Functions

FunctionWhat it does
synthesizeConverts text into spoken audio via ElevenLabs' documented Text to Speech API, returning it base64-encoded alongside the vendor's content type (MP3 by default). voiceId selects an…
cloneVoiceCreates a new ElevenLabs voice cloned from one or more caller-supplied audio sample URLs (Instant Voice Cloning) — downloads each sample and uploads it to ElevenLabs' voices/add. Returns…

Types

interface elevenlabsSynthesisResult {
  audioBase64: string;
  contentType: string;
  voiceId: string;
}
interface elevenlabsVoice {
  voiceId: string;
  name: string;
  requiresVerification: boolean;
}

Examples

const speech = await bowmark.providers.elevenlabs.synthesize({ text: "Hello there" });
// speech -> { audioBase64: "SUQzBAAAAAAAI1RTU0U…", contentType: "audio/mpeg", voiceId: "21m00Tcm4TlvDq8ikWAM" }
const voice = await bowmark.providers.elevenlabs.cloneVoice({
  name: "Narrator",
  sampleUrls: ["https://example.com/samples/narrator-01.mp3"],
});
// voice -> { voiceId: "c38kUX8pkfYO2kHyqfFy", name: "Narrator", requiresVerification: false }
const speech = await bowmark.providers.elevenlabs.synthesize({ text: "In a cloned voice.", voiceId: voice.voiceId });
text_to_speechText to speech — the capability this provider backs.