xAI TTS V1 turns a script into multilingual speech using one of five voices. Keep the words in the intended language and place pauses, laughter or passage-level whisper and slow-delivery tags where the performance should change. Choose the language or automatic detection, then configure codec and sample rate for playback, editing or a compatible telephony workflow. MP3 also supports a selectable bitrate. Review the trimmed script’s character estimate and audition the complete result before using the speech in a story, explanation or application.
xAI TTS V1 is developed by xAI. Vidgo provides access to the model. Model names and trademarks belong to their respective owners.Open the existing xAI speech task with the complete words to speak. Select eve, ara, rex, sal or leo, choose the language and add delivery tags in context. Configure the returned audio format and review the character estimate before generating.
The text field holds the words and inline cues, voice selects the speaker and language_code provides a supported language or auto. output_format selects codec, sample_rate and an MP3 bit_rate when applicable. Keep the source words in their actual spoken language when comparing them with the audio; page translations explain meaning without changing the recorded input.
| Model / task | Input and workflow |
|---|---|
| xAI TTS 1 Text to Speech | Expressive multilingual speech with five voices, inline speech tags, and configurable audio output. |
The existing recordings include an English museum story, a Mandarin kitchen-acoustics explanation and a Spanish neighborhood-repair announcement. Each uses its own text, voice and language with the recorded MP3 configuration. Listen to the separate results and inspect the full inputs to understand delivery cues and output settings.
Use tags such as [pause], [laugh] and a <whisper> passage beside the relevant words. The museum record uses eve to move from a discovery to a whispered detail and a comic realization. Listen for the changes in delivery and the continuity of the story.
Keep the explanation in its intended language and wrap the passage needing slower delivery in <slow> tags. The Mandarin acoustics record uses rex and zh, with a pause and a slower explanation of vibrating mass and frequency; review clarity and pacing in the audio.
Select a supported language code and one of the five voices, then write a complete announcement with appropriate pauses. The repair-workshop record uses ara and es-ES to invite neighbors to a shared activity, providing an actual source for reviewing Spanish speech.
The museum was closing when I spotted something beneath the dinosaur. [pause] <whisper>A tiny red mitten. With my name inside.</whisper> I had spent ten minutes looking for a priceless artifact. [laugh] Turns out, I was the mystery.
Write the exact script and place passage-level tags around the words they affect. The displayed museum text is the original English input, including its pause, whispered detail and laugh cue. Compare that script with the generated narration before adapting it.
{
"voice": "rex",
"language_code": "zh",
"output_format": {
"codec": "mp3",
"sample_rate": 24000,
"bit_rate": 128000
}
}Choose mp3, wav, pcm, mulaw or alaw and a supported sample rate. For MP3, select an allowed bit_rate; for the other codecs, configure codec and sample_rate. The shown request preserves its actual MP3, 24,000 Hz and 128,000-bit-per-second settings, which can be checked with the returned file.
The English museum request uses eve, a pause, a whispered passage and laughter to tell a missing-mitten story. The Mandarin kitchen request uses rex with a pause and a slower passage to explain why adding water changes a glass’s sound. Read the full original scripts, meaning translations and recorded voice/output settings while listening to each corresponding result.
The museum was closing when I spotted something beneath the dinosaur. [pause] <whisper>A tiny red mitten. With my name inside.</whisper> I had spent ten minutes looking for a priceless artifact. [laugh] Turns out, I was the mystery.
{
"voice": "eve",
"language_code": "en",
"output_format": {
"codec": "mp3",
"sample_rate": 24000,
"bit_rate": 128000
}
}Original Mandarin spoken text: 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。 English meaning: Tap an empty glass gently, then add a little water. Why does the sound become lower? [pause] Tapping makes the glass wall vibrate, and the water moves along with it. <slow>The mass vibrating together increases, so the vibration frequency decreases.</slow> Even a kitchen can hold a small acoustics laboratory.
{
"voice": "rex",
"language_code": "zh",
"output_format": {
"codec": "mp3",
"sample_rate": 24000,
"bit_rate": 128000
}
}Prepare the words in the intended language, choose a voice and place delivery cues where they support the message. Select the audio configuration, generate and review the complete result for story narration, educational explanations, public announcements, guided experiences, localized messages and application speech.
Write a complete short scene and place pauses or expressive passages where the story changes. Select a voice and listen for word clarity, delivery and the final reaction before using it in an audio narrative.
Prepare a clear concept explanation and use a slower passage where learners need time to follow it. Review pronunciation, pauses and pacing in the generated audio before placing it in a lesson.
Write the event, time, location and requested action in the intended language. Choose a voice and deliberate pauses, then review the complete spoken message for an announcement or shared activity.
Prepare concise spoken instructions with clear pauses and emphasis. Audition the chosen voice for ease of understanding before using the audio in a tour, onboarding sequence or guided task.
Prepare each message in its actual target language and select the supported language code. Generate and review pronunciation and delivery in each version before using the recordings in the localized experience.
Collect script, voice, language and output configuration, then submit the documented request. Retrieve the finished audio by task ID and use an encoding and sample rate compatible with the application’s playback or communication workflow.
xAI TTS uses the current rate per 1,000 Unicode characters after leading and trailing whitespace is trimmed. Inline tags, punctuation and spaces inside the script count toward the submitted text. Calculate the request amount proportionally from that count. The dynamic table shows the current rate; enter the full script in the task to review the estimate and plan additional recordings using their actual text lengths.
| Model | Provider | Workflow | Input material | Audio codecs | Starting rate |
|---|---|---|---|---|---|
| xAI TTS 1 Text to Speech | xAI | Text to Speech | Spoken text and delivery tags | MP3 / WAV / PCM / MULAW / ALAW | $0.012 / 1,000 characters |
Test your creative brief online, then bring the verified request into your application.
Open the xAI speech task and enter the full script with its delivery tags. Configure voice, language and output settings, review the character estimate and generate. Listen through the complete result for pronunciation, pauses and expression, then check the returned file with the planned playback or editing workflow.
Choose a task, inspect its settings and cost, and follow the result from an online test to an authenticated API request.
Prepare the words in the actual spoken language, choose eve, ara, rex, sal or leo and set a supported language code or auto. Place pauses and passage-level cues in context, then select codec, sample rate and an MP3 bitrate when applicable.
Open the xAI speech task and enter the full script with its delivery tags. Configure voice, language and output settings, review the character estimate and generate. Listen through the complete result for pronunciation, pauses and expression, then check the returned file with the planned playback or editing workflow.
Count the trimmed script’s Unicode characters, including all inline tags, internal spaces and punctuation. Use the task’s current per-1,000-character rate and the complete text for every planned take when reviewing the budget.
Read the original spoken text and voice settings while listening to the result. Translations explain the meaning while preserving the recording’s original spoken language. Review pronunciation, pacing and expression before adapting the request to your own script.
Retain the task ID returned by the API to query status and retrieve completed file URLs. Use your online generation history to review earlier results and the creative direction you tested.
Use the documentation linked from each task for its exact model identifier, accepted inputs, authentication and response format. Integrate the same task you tested online so its media and settings match the request.
xAI TTS V1 is a multilingual speech task on Vidgo with five voices, inline delivery tags and configurable audio encoding. It accepts up to 15,000 Unicode characters after trimming and provides language or automatic detection, codec, sample-rate and applicable MP3-bitrate settings.
Read speech-model introductions, API guides and platform comparisons to understand script preparation, voice controls, integration and character-based costs.
