Grok Speech·von xAISpeech to text
xAI STT v1
xAI STT v1 — transcribes an audio file to text. 24+ languages with automatic language detection, speaker diarization and filler-word control. The result is a downloadable text file.
xai/stt-v1Parameter
| Parameter | Typ | Standard | Bereich oder Optionen |
|---|---|---|---|
Raw audio format audio_formatFormat hint for raw/headerless audio. Container formats (mp3, wav, …) are auto-detected — leave on 'auto'. | select | auto | auto, pcm, mulaw, alaw |
Channels channelsNumber of audio channels (2–8). Only required for multichannel raw audio; auto-detected for container formats. | slider | 2 | 2 – 8 |
Speaker labels diarizeSpeaker diarization — each word carries the detected speaker id. | select | false | true, false |
Keep filler words filler_wordsWhen true, filler words ("um", "uh") are kept; when false they are removed. | select | false | true, false |
Language languageISO 639-1 language code (Filipino uses 'fil'). The model transcribes any supported language regardless; leave empty to auto-detect. | select | — | , ar, cs, da, nl, en, fil, fr, de, hi, id, it, ja, ko, mk, ms, fa, pl, pt, ro, ru, es, sv, th, tr, vi |
Per-channel transcript multichannelTranscribe each audio channel independently. | select | false | true, false |
Sample rate sample_rateSample rate in Hz. Only required for raw audio (pcm, mulaw, alaw). | select | 8000 | 8000, 16000, 22050, 24000, 44100, 48000 |
Number normalization text_normalizationConverts spoken-form numbers and currency into written form (e.g. "one hundred dollars" to "$100"). | select | false | true, false |
Eingaben
Dateien, die dieses Modell zusätzlich zum Prompt entgegennimmt.
| Eingabe | Akzeptiert | Max. Dateien | |
|---|---|---|---|
Audio File audio | audio | 1 | Erforderlich |