Fish Audio: Fish Transcribe-1

Fish Transcribe-1 converts an uploaded audio file into text with timed segments. It is useful for recorded speech, transcription workflows, and applications that need the transcript together with the audio's reported duration.

Compare models

fish-transcribe-1

AvailableFish AudioSpeech to TextSync

Price: From $0.0001
Modalities: Audio

About Fish Transcribe-1

Fish Transcribe-1 is Fish Audio's general speech-to-text model for short recordings. You upload an audio file and get back the transcript, the recording's duration and, when asked, word-level timed segments. Language is detected automatically, with an optional hint. It sits beside Transcribe-1 Pro, which Fish Audio recommends for long, multi-speaker conversations with emotion cues.

Where it works well

Responses include the audio's reported duration along with the text, which helps with billing checks and subtitle alignment.
Word-level start and end times are available when you request timestamps, so a transcript can be cut into captions.
The original file is sent as is: WAV, MP3, AAC, FLAC, Ogg, WebM/Matroska or MOV need no conversion first.
Language is detected automatically, so mixed batches of recordings do not need a language per file.

When to choose another model

Fish Audio positions this model for short recordings; multi-speaker or long audio is better served by Transcribe-1 Pro, with speaker markers and emotion cues.
It works on uploaded files, so live microphone streams need a realtime speech-to-text model.
Formats such as AIFF, CAF, WMA, AMR and raw PCM are not accepted and must be converted before upload.

Getting started

Create API key
Create a key in Console, then use it with every model on the platform.

Send your first request

Copy the example for your language and run it against the endpoint.

Speech to textTokenLab endpoint

POST/v1/audio/transcriptions

curl -X POST "https://api.tokenlab.sh/v1/audio/transcriptions" \
  -H "Authorization: Bearer sk-xxx" \
  -F file="@audio.mp3" \
  -F operation="stt" \
  -F response_format="json" \
  -F model="fish-transcribe-1" \
  -F language="en"

Pricing

The Verified price applies to Verified, which costs less on most models. The Official price is the model maker's published price and applies to the more reliable Official route. Auto bills the route that completes the request.

Speech to text

per second

Official price: $0.0001
Verified price: —
Discount: —

	Official priceper second	Verified priceper second	Discount
Speech to text	$0.0001	—	—

Usage & activity

Success rate is the share of requests that completed. Latency is how long a full response takes; P95 means 95% of requests finished within that time.

Usage & availability

Last 24 hours

Requests
Success rate
P95 latency
Total tokens

Model performance

Last updated Oct 14, 2026, 11:18 PM UTC

Metrics appear once privacy and data volume thresholds are met.

Data is based on aggregate user requests, excluding status checks.

Open in Console

Open Fish Transcribe-1 in Console with a prompt ready to edit or send.

Help me try fish-transcribe-1 with a short audio request at /v1/audio/transcriptions. Show the result and cost.

Use cases

Voice message transcription
Turn short voice notes, support call snippets and recorded memos into searchable text with the duration attached.
Caption drafts
Request word timings for a short clip and group them into subtitle lines for review.
Intake forms by voice
Transcribe spoken answers from a mobile app and pass the text on to a model that fills structured fields.
Multilingual archives
Run clips in different languages through one endpoint and rely on automatic language detection.

Prompt examples

Transcribe this 40-second voice memo and return word-level timestamps so I can build captions.

Transcribe the attached customer voicemail in its original language and give me the total duration of the recording.

Convert this MP3 of a short product demo narration to text; the speaker may switch between English and Mandarin.

FAQ

What is the difference between Fish Transcribe-1 and Transcribe-1 Pro?

Transcribe-1 is for general transcription of short recordings. Transcribe-1 Pro is the model Fish Audio recommends for recordings up to 60 minutes, multi-speaker conversations, and transcripts with speaker markers and emotion cues. Pick Pro when you need to know who said what.

What does Fish Transcribe-1 return?

The text, the duration in seconds, and a request ID. If you ask for timestamps it also returns word-level segments, each with text, start and end.

Which audio formats can I upload?

Fish Audio accepts WAV, MP3, AAC, FLAC, Ogg, WebM/Matroska and MOV files, sent as original bytes. For video containers it uses the first audio track. AIFF, CAF, WMA, AMR and raw PCM are not supported.

Do I need to tell it the language?

No. Fish Audio detects the language automatically, and the response includes the detected language when it is known. An optional language hint exists for cases where detection might be unsure, such as very short or noisy clips.

Can it transcribe live audio?

No. This model takes an uploaded file and returns the transcript after processing. For live captions or voice agents that must react while someone is speaking, choose a model built for streaming recognition.

How much does Fish Transcribe-1 cost?

On TokenLab, Fish Transcribe-1 costs $0.0001 per second. The pricing table above shows the full breakdown. Rates depend on the billing unit, specification, and usage. Compare matching conditions in the model's detailed pricing; a single rate does not determine the total cost.

Which endpoint should Fish Transcribe-1 use?

Use https://api.tokenlab.sh/v1/audio/transcriptions for Fish Transcribe-1. The request example below shows the matching code shape.

Which operations does Fish Transcribe-1 support?

Fish Transcribe-1 supports Speech to text. Select an operation above to see its endpoint and request example.