xAI TTS 1 Text to Speech API

xai/tts/v1

xAI TTS 1 Text to Speech turns scripts into multilingual audio with five voices and expressive speech tags. Place pauses and laughter within a line, or wrap a passage in tags to guide whispers and slower delivery.

Input

Text

232/15000

Voice

Language Code

Output Format

Output Format.Codec

Output Format.Sample Rate

Output Format.Bit Rate

Output

Speech ready

$0.012 / 1K chars · characters × 2.4 / 1000 credits · 0.5568 credits

Explore more audio tools

ElevenLabs V3 Text to Speech ElevenLabs Turbo v2.5 Text to Speech

Examples

A kitchen sound experiment.mp3

轻轻敲一下空玻璃杯，再往杯里加一点水，声音为什么变低了？[pause] 敲击让杯壁振动，水也跟着运动。<slow>一起振动的质量增加，振动频率就降低了。</slow> 厨房里，也藏着一间小小的声学实验室。

Repair it together.mp3

Este sábado abrimos el taller de reparación del barrio a las diez. [pause] Trae esa silla que cojea o una mochila con la cremallera rota. Compartiremos herramientas y aprenderemos a repararlas juntos. Nos vemos en la biblioteca.

xAI TTS 1 Text to Speech API

Generate speech with the API

Submit a script, keep the returned task_id, and retrieve the completed audio through the task status endpoint.

Authenticate your request

Send your Vidgo API key in the Authorization header using Bearer authentication.

Endpoint: POSThttps://api.vidgo.ai/api/generate/submit
Authentication: Authorization: Bearer YOUR_API_KEY

Submit the script

Send model and input as JSON. The example combines pauses and whispered speech.

Request example

REQUEST_BODY=$(cat <<'JSON'
{
  "model": "xai/tts/v1",
  "input": {
    "text": "Welcome to our story. [pause] <whisper>Listen closely.</whisper> [laugh] Let us begin.",
    "voice": "eve",
    "language_code": "auto",
    "output_format": {
      "codec": "mp3",
      "sample_rate": 24000,
      "bit_rate": 128000
    }
  }
}
JSON
)

RESPONSE=$(curl --silent --show-error --fail-with-body \
  --request POST \
  --url "https://api.vidgo.ai/api/generate/submit" \
  --header "Authorization: Bearer $VIDGO_API_KEY" \
  --header "Content-Type: application/json" \
  --data "$REQUEST_BODY")

CODE=$(printf '%s' "$RESPONSE" | jq -r '.code // empty')
if [ "$CODE" != "0" ] && [ "$CODE" != "200" ]; then
  printf 'API error: %s
' "$RESPONSE" >&2
  exit 1
fi
printf '%s
' "$RESPONSE"

Follow the task

Query the returned task_id until status is finished or failed.

Status endpoint

GET https://api.vidgo.ai/api/generate/status/{task_id}

Keep the task_id from submission. Check its status and read the generated files when it finishes.

not_startedrunningfinishedfailed

Submission response

{
  "code": 200,
  "data": {
    "task_id": "your_task_id",
    "status": "not_started",
    "created_time": "2026-09-27T10:00:00"
  }
}

Completed response

{
  "code": 200,
  "data": {
    "task_id": "your_task_id",
    "status": "finished",
    "files": [
      {
        "file_url": "https://example.com/generated-speech.mp3",
        "file_type": "audio",
        "file_name": "speech.mp3"
      }
    ],
    "error_message": null,
    "progress": 100
  }
}

Read the audio files

Read files with file_type set to audio, then use file_url to download the speech.

Read the audio files

set -euo pipefail
: "${VIDGO_API_KEY:?Set VIDGO_API_KEY in your environment}"

REQUEST_BODY=$(cat <<'JSON'
{
  "model": "xai/tts/v1",
  "input": {
    "text": "Welcome to our story. [pause] <whisper>Listen closely.</whisper> [laugh] Let us begin.",
    "voice": "eve",
    "language_code": "auto",
    "output_format": {
      "codec": "mp3",
      "sample_rate": 24000,
      "bit_rate": 128000
    }
  }
}
JSON
)

SUBMIT_RESPONSE=$(curl --silent --show-error --fail-with-body \
  --request POST \
  --url "https://api.vidgo.ai/api/generate/submit" \
  --header "Authorization: Bearer $VIDGO_API_KEY" \
  --header "Content-Type: application/json" \
  --data "$REQUEST_BODY")

TASK_ID=$(printf '%s' "$SUBMIT_RESPONSE" | jq -r '.data.task_id // .task_id // empty')
BUSINESS_CODE=$(printf '%s' "$SUBMIT_RESPONSE" | jq -r '.code // empty')
if [ "$BUSINESS_CODE" != "0" ] && [ "$BUSINESS_CODE" != "200" ]; then
  printf 'Submit failed:
%s
' "$SUBMIT_RESPONSE" >&2
  exit 1
fi
if [ -z "$TASK_ID" ]; then
  printf 'Submit response did not include task_id:
%s
' "$SUBMIT_RESPONSE" >&2
  exit 1
fi

START_TIME=$(date +%s)
POLL_DELAY=2
while true; do
  if [ $(( $(date +%s) - START_TIME )) -ge 600 ]; then
    printf 'Timed out after 600 seconds
' >&2
    exit 1
  fi
  STATUS_RESPONSE=$(curl --silent --show-error --fail-with-body \
    --url "https://api.vidgo.ai/api/generate/status/$TASK_ID" \
    --header "Authorization: Bearer $VIDGO_API_KEY")
  STATUS=$(printf '%s' "$STATUS_RESPONSE" | jq -r '.data.status // .status // empty')
  BUSINESS_CODE=$(printf '%s' "$STATUS_RESPONSE" | jq -r '.code // empty')
  if [ "$BUSINESS_CODE" != "0" ] && [ "$BUSINESS_CODE" != "200" ]; then
    printf 'Status request failed:
%s
' "$STATUS_RESPONSE" >&2
    exit 1
  fi

  case "$STATUS" in
    finished)
      printf '%s' "$STATUS_RESPONSE" | jq -r '(.data.files // .files // [])[]?.file_url'
      break
      ;;
    failed)
      printf '%s' "$STATUS_RESPONSE" | jq -r '.data.error_message // .error_message // "Generation failed"' >&2
      exit 1
      ;;
    not_started|running)
      sleep "$POLL_DELAY"
      if [ "$POLL_DELAY" -lt 10 ]; then POLL_DELAY=$((POLL_DELAY + 1)); fi
      ;;
    *)
      printf 'Unexpected task status: %s
' "$STATUS" >&2
      exit 1
      ;;
  esac
done

Request parameters

Place speech controls inside input; keep model and callback_url at the top level.

Field	Type	Required	Default	Description
model	string	Yes	—	Set model to xai/tts/v1.
input	object	Yes	—	Contains the script and speech controls listed below.
input.text	string	Yes	—	Text to speak, from 1 to 15,000 characters after trimming leading and trailing whitespace. Use [laugh], [pause], [sigh], <whisper>text</whisper>, or <slow>text</slow> to direct delivery.
input.voice	string	No	eve	Choose eve, ara, rex, sal, or leo for your narrator or character. Default: eve.
input.language_code	string	No	auto	Choose a listed BCP-47 language code, or auto for automatic language detection. auto, en, ar-EG, ar-SA, ar-AE, bn, zh, fr, de, hi, id, it, ja, ko, pt-BR, pt-PT, ru, es-MX, es-ES, tr, vi
input.output_format	object	No	—	An object containing codec, sample_rate, and the MP3 bit_rate setting.
input.output_format.codec	string	No	mp3	Audio encoding: mp3, wav, pcm, mulaw, or alaw. Default: mp3.
input.output_format.sample_rate	integer	No	24000	Sample rate in Hz: 8000, 16000, 22050, 24000, 44100, or 48000. Default: 24000.
input.output_format.bit_rate	integer \| null	No	128000 (mp3)	MP3 bitrate in bits per second: 32000, 64000, 96000, 128000, 192000, or null. MP3 default: 128000. For wav, pcm, mulaw, and alaw, omit bit_rate or set it to null.
callback_url	string	No	—	Optional HTTP or HTTPS URL for task notifications. Include it at the top level of the request.

Response fields

Submission identifies the task. Status responses report progress toward the completed audio.

Field	Type	Description
code	integer	Response code. A successful response uses 200 or 0.
data.task_id	string	Task identifier used to query status.
data.status	string	Task state: not_started, running, finished, or failed.
data.created_time	string	Task creation time returned by the service.
data.files	array	Generated speech files.
data.files[].file_url	string	URL used to play or download a generated file.
data.files[].file_type	string	Generated speech uses audio.
data.files[].file_name	string	File name, when returned.
data.error_message	string \| null	Error details when a task fails.

Task lifecycle

Use the task state to decide when to check again or read the result.

not_started
The task has been accepted and is waiting to start.
running
Speech generation is in progress. Continue checking the same task.
finished
Generation is complete. Read files to retrieve the speech.
failed
Generation failed. Read error_message before deciding whether to submit a new request.

Handle results and interruptions

Keep the current task identifierIf a status request is interrupted, query the same task_id again to recover its latest state.
Read the audio resultUse the file_url and file_name from each audio entry to download the speech.
Inspect a failed taskWhen status is failed, read error_message and review the script and settings before submitting another generation.

Speech specifications

Item	Value	Description
Task	Text to Speech	Generate speech from text and a selected voice.
Text length	1–15,000	Unicode characters after trimming leading and trailing whitespace.
Voices	eve, ara, rex, sal, leo	Select the preset through input.voice.
Languages	auto, en, ar-EG, ar-SA, ar-AE, bn, zh, fr, de, hi, id, it, ja, ko, pt-BR, pt-PT, ru, es-MX, es-ES, tr, vi	Choose a language code or auto.
Output	mp3, wav, pcm, mulaw, alaw	Configure the encoding and sample rate in output_format.

View the xAI TTS 1 Text to Speech API docs

xAI TTS 1 Text to Speech

xAI TTS 1 Text to Speech is xAI’s speech generation model for multilingual narration and character lines. Inline tags guide pauses, laughter and sighs, while wrapping tags direct whispers or slower delivery across a passage. Choose a voice, language, codec and sample rate to prepare audio for video editing, course production or integration with audio systems.

Why Choose xAI TTS 1

Direct expression within the scriptUse inline tags for pauses, laughter or sighs and wrapping tags for whispered or slower passages, keeping delivery directions beside the relevant words.
Choose character voices from five presetsCompare eve, ara, rex, sal and leo with the same lines, choose a voice for a narrator or character, and use that preset for subsequent passages.
Prepare speech for multilingual contentDetect the language from the script or select a language and regional code to voice localized lessons, product explanations and stories.
Prepare audio for editing and integrationUse MP3 for playback, WAV for editing, or PCM, mulaw and alaw for audio systems, then set the sample rate and MP3 bitrate to match the destination.

Speech parameters

Parameter	Requirement	What it controls
text	Required	Text to speak, from 1 to 15,000 characters after trimming leading and trailing whitespace. Use [laugh], [pause], [sigh], <whisper>text</whisper>, or <slow>text</slow> to direct delivery.
voice	Optional	Choose eve, ara, rex, sal, or leo for your narrator or character. Default: eve. Default`eveararexsalleo`
language_code	Optional	Choose a listed BCP-47 language code, or auto for automatic language detection. Default`autoenar-EGar-SAar-AEbnzhfrdehiiditjakopt-BRpt-PTrues-MXes-EStrvi`
output_format	Optional	An object containing codec, sample_rate, and the MP3 bit_rate setting.
output_format.codec	Optional	Audio encoding: mp3, wav, pcm, mulaw, or alaw. Default: mp3. Default`mp3wavpcmmulawalaw`
output_format.sample_rate	Optional	Sample rate in Hz: 8000, 16000, 22050, 24000, 44100, or 48000. Default: 24000. Default`24000800016000220504410048000`
output_format.bit_rate	Optional	MP3 bitrate in bits per second: 32000, 64000, 96000, 128000, 192000, or null. MP3 default: 128000. For wav, pcm, mulaw, and alaw, omit bit_rate or set it to null. Default`128000 (mp3)320006400096000128000192000null`

Create your first voiceover

Prepare the spoken scriptEnter the script in Text, add [pause] or [laugh] where needed, and use paired tags around whispered or slower passages.
Choose a voice and languageChoose a voice in Voice and a language code or auto in Language Code. Start with a short passage to review the delivery.
Set output and submitChoose Output Format.Codec and Output Format.Sample Rate. For mp3, set Output Format.Bit Rate, review the estimate and select Run.
Review and save the audioPreview MP3 and WAV in the output panel. Download any format, including PCM, mulaw and alaw for use in an audio application.

Character-based pricing

xAI TTS 1 Text to Speech costs $0.012, or 2.4 credits, per 1,000 characters. Usage is proportional to the Unicode character count after trimming leading and trailing whitespace. Speech tags, punctuation, and spaces within the text count toward usage.

Text length	Cost	Calculation
1 characters	$0.000012 · 0.0024 credits	1 × 2.4 / 1000 = 0.0024 credits
1,000 characters	$0.012 · 2.4 credits	1000 × 2.4 / 1000 = 2.4 credits
15,000 characters	$0.18 · 36 credits	15000 × 2.4 / 1000 = 36 credits

Put speech to work

Story narrationMark pauses at narrative turns and use whispered passages for quiet lines in short films and spoken stories.
Character linesChoose voices for characters and generate lines with laughter, sighs or changes in pace, then assemble them during editing.
Localized coursesChoose a language for translated lesson scripts and use the selected voice to produce terminology explanations and instructional audio.

Shape the delivery

Place [pause] at sentence boundaries and [laugh] or [sigh] where a line calls for that expression.
Use complete pairs such as <whisper>words</whisper> or <slow>words</slow> around the passage you want to direct.
Compare the five voices with the same short passage, then keep the chosen voice and language for the full narration.
Split long scripts at paragraph boundaries, keep each request within 15,000 characters and retain the same codec and sample rate for editing.

Choose output settings

The default output is MP3 at 24000 Hz and 128000 bit/s.
For wav, pcm, mulaw, or alaw requests, provide codec and sample_rate; omit bit_rate or set it to null.

Explore more audio models

ElevenLabs V3 Text to SpeechExplore expressive speech with ElevenLabs.

ElevenLabs Turbo v2.5 Text to SpeechExplore another speech model.

xAI TTS 1 Text to Speech API frequently asked questions

What is the xAI TTS 1 Text to Speech API?

xAI TTS 1 Text to Speech is an xAI model that turns scripts into narration and character lines. It offers five voices, multilingual speech and expressive text tags for stories and lessons. Inline tags direct local pauses and vocal expressions, while wrapping tags guide whispers or slower delivery across a passage. You can call it programmatically or try it in the Playground tab.

Which voices can I choose in xAI TTS 1?

Voice offers eve, ara, rex, sal and leo, with eve as the default. Compare them using the same lines before choosing a voice for narration or a character.

How do inline speech tags work in xAI TTS 1?

Add [pause], [laugh] or [sigh] in Text where you want a pause, laugh or sigh. Keep tags next to the relevant words so you can review and refine delivery line by line.

How do wrapping tags work in xAI TTS 1?

Use <whisper>words</whisper> to guide whispers and <slow>words</slow> for slower delivery. Paired tags enclose the text you want to direct, allowing different passages within one script to use different delivery styles.

How do I choose a language in xAI TTS 1?

Set Language Code to auto for automatic detection, or select a code such as zh, en or ja. Portuguese, Spanish and Arabic have regional options in the list; choose the code that matches the script.

How long can an xAI TTS 1 script be?

Each request accepts 1–15,000 Unicode characters after trimming leading and trailing whitespace. Tags, punctuation and internal spaces count toward the limit. Split longer work into passages using the same voice, language and output settings.

Which audio formats does xAI TTS 1 offer?

Output Format.Codec offers mp3, wav, pcm, mulaw and alaw. Preview MP3 and WAV on the page, or download PCM, mulaw and alaw for audio systems. Choose a sample rate from the options listed in Parameters.

How do I set the MP3 bitrate in xAI TTS 1?

For mp3, choose 32000, 64000, 96000, 128000 or 192000 in Output Format.Bit Rate, measured in bit/s; the default is 128000. With other codecs, omit bit_rate or set it to null and configure the codec and sample rate.