OpenAI SDK-compatible speech

POST/oai/audio/speech

Generates speech with the official OpenAI SDK. Set the client's base URL to https://api.allmodels.io/oai and use your AllModels API key. Use stream_format=audio for audio bytes or stream_format=sse for OpenAI speech events.

Use provider_options for provider-specific settings. You can also pass supported options directly as query parameters. If both forms set the same option, provider_options[<name>]=<value> takes precedence. Recognized enum and boolean option values are case-insensitive and are normalized to each provider's wire spelling; free-form values such as prompts, keyterms, and voice IDs retain their original case.

Authorization

BearerAuth

AuthorizationBearer <token>

Tenant key supplied as the Authorization Bearer token.

In: header

Query Parameters

provider_order?array<string>

Comma-separated or repeated provider IDs in preferred order. Providers not listed remain eligible.

provider_only?array<string>

Comma-separated or repeated provider IDs. Only these providers may serve the request.

provider_ignore?array<string>

Comma-separated or repeated provider IDs that must not serve the request.

allow_fallbacks?boolean

Set to false to prevent fallback to another provider. Automatic provider retries are not currently supported.

provider_options?||||||||||||||

Provider-specific options in bracket notation, such as provider_options[encoding]=linear16. You can also pass supported options directly as query parameters. If both forms set the same option, the bracketed value takes precedence.

Request Body

application/json

TypeScript Definitions

Use the request body type in TypeScript.

input*string

Text to synthesize.

Length1 <= length <= 4096

model*string

Model id, optionally provider-prefixed.

Length1 <= length

voice*|

Voice id string, or an object carrying it as id. A voice you cloned with AllModels is named by its voice ID, e.g. AM:@narrator.

instructions?string

Free-form delivery instructions (tone, accent, pacing, character). Forwarded natively where the selected binding has a dedicated instruction field, and otherwise rendered into the mechanism that binding documents; the per-binding instructions capability published on /v1/models says which. A binding whose mechanism is a finite tag vocabulary (mode: inline-fixed) honors per request: it renders an instruction its live-verified vocabulary covers whole, and refuses one it does not rather than rendering part of it, so the same wording can be honored on one binding and refused on another. Honoring can also be placement-dependent: a binding whose tags render measurably weaker applied per streamed text frame honors on this synchronous route and refuses on the streaming TTS socket. A binding that cannot honor the parameter is never served a silently downgraded rendition — the request returns 422 unsupported_parameter instead. Maximum 4096 characters.

Lengthlength <= 4096

response_format?string

Output format: one of mp3, opus, aac, flac, wav, pcm (case-insensitive). Formats the selected provider cannot emit natively are served by transcoding its best native stream (audio responses only, not stream_format=sse); combinations that cannot be transcoded return invalid_response_format.

speed?number

Playback speed, where supported.

stream_format?string

Value in

"audio"
"sse"

provider?unknown

Provider preferences.

provider_options?||||||||||||||

[key: string]?unknown

Response Body

`application/json`

import { writeFile } from "node:fs/promises";import OpenAI from "openai";const client = new OpenAI({  apiKey: process.env.ALLMODELS_API_KEY,  baseURL: "https://api.allmodels.io/oai"});const response = await client.audio.speech.create({  model: "elevenlabs/eleven-turbo-v2-5",  voice: "pNInz6obpgDQGcFmaJgB",  input: "Hello from AllModels",  response_format: "mp3"});await writeFile("speech.mp3", Buffer.from(await response.arrayBuffer()));

import osfrom openai import OpenAIclient = OpenAI(    api_key=os.environ["ALLMODELS_API_KEY"],    base_url="https://api.allmodels.io/oai",)with client.audio.speech.with_streaming_response.create(    model="elevenlabs/eleven-turbo-v2-5",    voice="pNInz6obpgDQGcFmaJgB",    input="Hello from AllModels",    response_format="mp3",) as response:    response.stream_to_file("speech.mp3")

"string"

{  "error": {    "message": "string",    "type": "string",    "param": "string",    "code": "string"  }}

{  "error": {    "message": "string",    "type": "string",    "param": "string",    "code": "string"  }}

{  "error": {    "message": "string",    "type": "string",    "param": "string",    "code": "string"  }}

{  "error": {    "message": "string",    "type": "string",    "param": "string",    "code": "string"  }}

{  "error": {    "message": "string",    "type": "string",    "param": "string",    "code": "string"  }}

{  "error": {    "message": "string",    "type": "string",    "param": "string",    "code": "string"  }}

Authorization

Query Parameters

Request Body

Response Body

200application/octet-stream

400application/json

401application/json

402application/json

403application/json

422application/json

502application/json

`application/json`

`application/json`

`application/json`

`application/json`

`application/json`

`application/json`