AbhiBots/ docs
Back to Docs

TTS API

Convert text to natural-sounding speech

Overview

The TTS API allows you to convert text into high-quality audio using various AI voice providers including ElevenLabs, OpenAI, Google, and Gemini. Each character of text costs 1 credit. The response is the generated audio itself, streamed back as an audio/mpeg (or audio/wav) file.

💰 Pricing: 1 character = 1 credit

POST/api/tts/generate

Generate audio from text. Returns audio file or base64 encoded audio.

Request Body

ParameterTypeRequiredDescription
textstringYesText to convert (max 5000 chars)
voiceIdstringYesVoice ID from voices endpoint
providerstringNoProvider: eleven, openai, google, gemini (default: eleven)
voice_settingsobjectNoElevenLabs tuning object: model_id (see models below), stability 0-1, similarity 0-1, style 0-1, speaker_boost.
sourcestringNoElevenLabs voice engine: c, f or p. Omit to let the system auto-select the best available (recommended).

ElevenLabs Models — voice_settings.model_id

Choose how your ElevenLabs voice is rendered by passing model_id inside voice_settings. Default is eleven_flash_v2_5. Longer text is automatically split into chunks and stitched into one complete file — you never get a cut-off clip.

model_idBest for
eleven_flash_v2_5Fastest, lowest latency — ideal for long narration (default)
eleven_turbo_v2_5Fast with strong quality — balanced choice
eleven_multilingual_v2Highest quality, 29 languages
eleven_v4_turboNewest v4 Turbo — fast & high quality (official route, 4x credits)
eleven_v3Most expressive — supports emotion audio-tags like [excited], [whispers], [sighs]
Voice vs model: the voiceId decides who speaks; the model_id decides how. You always get exactly the voice you selected — the model only changes speed/expressiveness/quality.

Example Request

curl -X POST "https://app.abhibots.com/api/tts/generate" \
  -H "X-API-Key: your_api_key" \
  -H "Content-Type: application/json" \
  -d '{
    "text": "Hello! Welcome to AbhiBots TTS API.",
    "voiceId": "21m00Tcm4TlvDq8ikWAM",
    "provider": "eleven",
    "voice_settings": {
      "model_id": "eleven_flash_v2_5",
      "stability": 0.5,
      "similarity": 0.75
    }
  }'

Response

On success (200) the endpoint returns the audio file directly as a binary stream — Content-Type: audio/mpeg for ElevenLabs/OpenAI or audio/wav for Gemini. Write the response body to a file. On failure it returns JSON { "detail": "..." } with a 4xx/5xx status (e.g. 401 bad key, 402 no credits, 429 rate limit).

# Save the audio straight to a file
curl -X POST "https://app.abhibots.com/api/tts/generate" \
  -H "X-API-Key: your_api_key" -H "Content-Type: application/json" \
  -d '{"text":"Hi there","voiceId":"21m00Tcm4TlvDq8ikWAM","provider":"eleven","voice_settings":{"model_id":"eleven_flash_v2_5"}}' \
  --output speech.mp3

Python Example

import requests

API_KEY = "your_api_key"
API_URL = "https://app.abhibots.com/api/tts/generate"

response = requests.post(
    API_URL,
    headers={"X-API-Key": API_KEY},
    json={
        "text": "Hello from Python!",
        "voiceId": "21m00Tcm4TlvDq8ikWAM",
        "provider": "eleven",
        "voice_settings": {"model_id": "eleven_flash_v2_5"}
    }
)

if response.status_code == 200:
    data = response.json()
    print(f"Audio URL: {data['audio_url']}")
    print(f"Credits used: {data['credits_used']}")
else:
    print(f"Error: {response.json()}")

JavaScript Example

const API_KEY = "your_api_key";

const response = await fetch("https://app.abhibots.com/api/tts/generate", {
    method: "POST",
    headers: {
        "X-API-Key": API_KEY,
        "Content-Type": "application/json"
    },
    body: JSON.stringify({
        text: "Hello from JavaScript!",
        voiceId: "21m00Tcm4TlvDq8ikWAM",
        provider: "eleven",
        voice_settings: { model_id: "eleven_flash_v2_5" }
    })
});

const data = await response.json();
console.log("Audio URL:", data.audio_url);
console.log("Credits used:", data.credits_used);
POST/api/tts/streamREAL-TIME

Real-time streaming TTS via Gemini Live API (WebSocket-backed). First audio chunk arrives in ~1 second — play progressively before the full clip is ready. Charged 2x (Gemini). Returns a live audio/wav stream.

Request Body

ParameterTypeRequiredDescription
textstringYesText to convert (up to 20,000 chars)
voiceIdstringNoGemini voice (default Kore): Aoede, Puck, Zephyr
modelstringNoflash (fast) or pro (quality). Default flash
stylestringNoStyle prompt e.g. cheerful, calm

Example (curl)

curl -X POST "https://app.abhibots.com/api/tts/stream" -H "X-API-Key: your_api_key" -H "Content-Type: application/json" -d '{"text":"Hello in real time!","voiceId":"Kore","model":"flash"}' --output stream.wav

Python (stream to file)

import requests

with requests.post(
    "https://app.abhibots.com/api/tts/stream",
    headers={"X-API-Key": "your_api_key"},
    json={"text": "Streaming!", "voiceId": "Kore", "model": "flash"},
    stream=True) as r:
    with open("out.wav", "wb") as f:
        for chunk in r.iter_content(8192):
            f.write(chunk)

Real-time playback (play as it streams)

# Play in real-time as chunks arrive (needs: pip install pyaudio)
import requests, pyaudio

pa = pyaudio.PyAudio()
stream = pa.open(format=pyaudio.paInt16, channels=1, rate=24000, output=True)

with requests.post(
    "https://app.abhibots.com/api/tts/stream",
    headers={"X-API-Key": "your_api_key"},
    json={"text": "Plays as it streams!", "voiceId": "Kore"},
    stream=True) as r:
    first = True
    for chunk in r.iter_content(4096):
        if first:  # skip 44-byte WAV header
            chunk = chunk[44:]; first = False
        stream.write(chunk)
Streaming uses Gemini voices only. For ElevenLabs voices use /api/tts/generate or /api/tts/task.
POST/api/tts/taskASYNC

Best for long text and SaaS / batch use. Returns a task_id instantly (no waiting), then you poll the Get Task endpoint below for the result. Avoids sync-request timeouts on big generations. ElevenLabs voices. Note: the text field is input.

Request Body

ParameterTypeRequiredDescription
inputstringYesText to convert (up to 50,000 chars)
voice_idstringYesVoice ID from the voices endpoint
model_idstringNoeleven_flash_v2_5 (default), eleven_multilingual_v2, eleven_v3, eleven_turbo_v2_5, eleven_v4_turbo
stabilityfloatNo0-1 (default 0.5). Premium tuning
similarityfloatNo0-1 (default 0.75). Premium tuning
sourcestringNoOptional engine hint: c, f or p

Example (curl)

curl -X POST "https://app.abhibots.com/api/tts/task" -H "X-API-Key: your_api_key" -H "Content-Type: application/json" -d '{"input":"Long text here...","voice_id":"21m00Tcm4TlvDq8ikWAM","model_id":"eleven_flash_v2_5"}'

Response

{ "task_id": "a1b2c3d4" }
GET/api/tts/task/{task_id}

Poll this until status is completed, then download the audio from the result URL.

Response (completed)

{
  "id": "a1b2c3d4",
  "status": "completed",        // processing | completed | failed
  "result": "https://app.abhibots.com/api/tts/audio/12345_abcd.mp3",
  "input": "Long text here...",
  "voice_id": "21m00Tcm4TlvDq8ikWAM",
  "model_id": "eleven_flash_v2_5"
}

Python (create -> poll -> download)

import requests, time

H = {"X-API-Key": "your_api_key"}
task = requests.post("https://app.abhibots.com/api/tts/task", headers=H,
    json={"input": "Long text...", "voice_id": "21m00Tcm4TlvDq8ikWAM"}).json()
tid = task["task_id"]

while True:
    t = requests.get(f"https://app.abhibots.com/api/tts/task/{tid}", headers=H).json()
    if t["status"] == "completed":
        audio = requests.get(t["result"], headers=H).content
        open("out.mp3", "wb").write(audio); break
    if t["status"] == "failed":
        raise Exception(t.get("error"))
    time.sleep(2)
GET/api/tts/history

Returns your account's most recent generations (up to 50, newest first) - text, voice, characters, duration and a downloadable audio URL.

Example (curl)

curl "https://app.abhibots.com/api/tts/history" -H "X-API-Key: your_api_key"

Response

[
  {
    "id": "344079",
    "provider": "eleven",
    "voice": "21m00Tcm4TlvDq8ikWAM",
    "text": "First chars of the text...",
    "chars": 1483,
    "duration": 98.8,
    "createdAt": "2026-08-22T13:27:31Z",
    "audioUrl": "https://app.abhibots.com/api/tts/audio/12345_abcd.mp3"
  }
]
Built by · Founder of AbhiBots