GPT Transcribe - Kyma API

Overview

GPT Transcribe converts recorded audio into text. It targets the hard cases rather than clean studio audio: conversational speech, several speakers, background noise, and sentences that switch language mid-way (Vietnamese and English in one pass, for example). On Kyma it is served through the same OpenAI-compatible transcription endpoint as the other speech-to-text SKUs — multipart file or JSON audio_url, with json and text response formats. Billing is per minute of audio, rounded up, with a one-minute minimum; the exact figure comes back in the response and in the X-Kyma-Cost-USD header. This is file transcription. It is not the live streaming surface, and it is not audio-scene understanding. It returns a transcript, not segment timestamps, so subtitle work stays on whisper-v3-turbo.

Specs

Field	Value
Model ID	`gpt-transcribe`
Best for	High-accuracy dictation, multilingual transcription, conversational audio
Context window	1,500
Input modalities	Audio
Output modalities	Text
Speed	Fast
Cost band	Balanced
Release stage	Stable

Pricing

	Price
Per minute of audio	$0.006075

Billed per minute of audio. Every request reports its exact cost in usage.cost; the full media price list is on Pricing.

Use this when

Conversational audio transcription — Captures multi-speaker dialogue and natural speech patterns with high accuracy.
Noisy background dictation — Filters background interference to produce clean text from field recordings or calls.
Mixed language audio processing — Handles code-switched audio streams like Vietnamese and English in a single pass.
Migration off the gpt-4o-transcribe family — The id OpenAI names as the replacement for the models it shuts down on 2027-02-26.

Not ideal for

Do not use this model for generating audio, for live bidirectional speech, or when you need segment timestamps — it does not return them, and SRT/VTT output needs a model that does.

Pick something else when

You need segment timestamps, SRT or VTT output, or the cheapest per-minute rate on clear recordings: use whisper-v3-turbo.
You want a file-transcription SKU from a different creator: use gemini-3.5-transcribe.

Example

from openai import OpenAI

client = OpenAI(base_url="https://kymaapi.com/v1", api_key="kyma-...")

response = client.audio.transcriptions.create(
    model="gpt-transcribe",
    file=open("audio.wav", "rb"),
)

Agent query example

Ask the API which models fit, instead of hardcoding an id:

curl "https://kymaapi.com/v1/models?recommended_for=transcribe-audio&quality_tier=frontier-open"

FAQ

Does the transcribe-quality alias point here? Not yet. transcribe-quality still resolves to gpt-4o-mini-transcribe-2025-12-15. Pin gpt-transcribe by id when you want this model. Can I get SRT or VTT subtitles from this model? No. It returns a transcript without segment timestamps, so the subtitle formats are not available on this id. Use whisper-v3-turbo for timestamped output. How is this billed? Per minute of audio, rounded up, with a one-minute minimum. The Pricing table on this page is the live rate.

​Overview

​Specs

​Pricing

​Use this when

​Not ideal for

​Pick something else when

​Example

​Agent query example

​FAQ