Tasks and inputs
Transcribe 1 Pro is listed for Automatic speech recognition, Speech → Text, Speaker Diarization. It accepts audio, video, image and produces text. Use the publisher’s input format and preprocessing for your chosen deployment.
SyncDev model directory · Profile updated Nov 17, 2026
Fish Audio Transcribe 1 Pro is a multi-speaker speech-to-text model with inline speaker labels, emotion cues and word timestamps, for files up to 60 minutes.
The specification score summarizes recorded features and limits. It is not a benchmark result.
Pricing conditions: Indicative: $0.0001 per audio second (~$0.36 per audio hour). Confirm on your account.
Recorded modality support shows what this model can accept and produce. An undocumented modality is shown as unknown rather than assumed to be unsupported.
Accepts
Produces
Here is a trap worth knowing before you write a line of code. Fish Audio Transcribe 1 Pro is selected with a model header on POST /v1/asr. If that header is missing or mistyped, the request does not fail. It silently falls back to the older transcribe-1, and bills as that model. Validate the header in your client, and check which model your invoices show.
With that out of the way: Transcribe 1 Pro is Fish Audio's recommended speech-recognition model, built for recordings with several people talking.
The transcript carries more than words.
That last default surprises people. If you are feeding transcripts to a search index or a summariser, the cues may be noise; turn them off. For podcast show notes or call QA, they are often exactly what you want.
| Item | Detail |
|---|---|
| API ID | transcribe-1-pro |
| Endpoint | POST /v1/asr with model header |
| Max length | 60 minutes |
| Formats | WAV, MP3, AAC/M4A, FLAC, Ogg, WebM/Matroska, MOV |
| Video files | First audio track is used |
| Indicative price | $0.0001 per audio second ($0.36 per hour) |
For a recording with commentary on a second audio track, extract the track you want first, since only the first one is read.
Fish's documentation gives no release date. A dated listing slug says 24 September 2026; a registry says 18 September; coverage appeared on 28 September. We use 24 September as indicative. Coverage also cites 83 supported languages, but Fish's speech-to-text docs publish no language list, so test your languages rather than assuming.
The older transcribe-1 is deprecated and retires on 31 December 2026. Given the fallback behaviour above, any integration that has not been updated will break at the end of the year. Search your codebase now.
Speaker labelling is where transcription models differ most. Take five recordings with overlapping speech, three or more voices and at least one long monologue. Mark the true speaker turns by hand for five minutes of each, then compare. Count speaker-swap errors, not just word errors. For a wider set of options see speaker diarization models and speech recognition models.
Compare the model on the work you need to complete.
Transcribe 1 Pro is listed for Automatic speech recognition, Speech → Text, Speaker Diarization. It accepts audio, video, image and produces text. Use the publisher’s input format and preprocessing for your chosen deployment.
Measure a completed task using the prompts, documents and tools your application will use. Include reasoning and retries in the result.
Use the model identifier and account requirements for the selected API route. Confirm its request limits and output format with one complete task before connecting a larger workflow.
These records share a directory category. Check the source card, output type and deployment requirements before treating a model as a replacement.
ElevenLabs
Scribe v3 Live is ElevenLabs' streaming transcription model with turn prediction, mid-sentence language switching and a 10,000-term keyterm library.
Microsoft
MAI-Transcribe-2-Streaming transcribes live audio with intermediate and final segments through Microsoft Foundry’s speech services.
assemblyai
AssemblyAI Universal-3.6 Pro is a realtime streaming speech-to-text model covering 32 languages with code switching, at $0.45 per audio hour.
xAI
Grok Voice Transcribe 2.0 transcribes recorded and streaming audio; the retired 1.0 API name now routes to this revision.
Gemini 3.5 Transcribe Live from Google: Speech transcription model for accurate audio-to-text and captioning workflows.
xAI
Grok Voice STT 1.0 from xAI: Grok Voice STT 1.0 is xAI's speech-to-text model. It supports transcription with word-level timestamps, optional speaker diarization, and multichannel audio.