OpenAI: GPT Transcribe

Overview

Speech-to-text model for audio files and committed turns in Realtime sessions, with support for context and keyword hints.

Provider

OpenAI

Modality

Transcription

Input

audio, text

Output

text

Weights

Closed

Distillable

NoLicence or terms, checked October 8, 2026.

Tags

speech-to-text, audio

Docs

Provider documentation

Data policy

Prompts go to

OpenAI (no processing-region constraint by default)

Retention

API data not used for training unless you opt in, abuse monitoring logs that can include prompts and responses kept up to 30 days by default, some features store application state

Zero data retention

Available

Region

Global by default, data residency projects in US, EU and other regions (approval required, 10% uplift for models released on or after 2026-03-05)

Policy

OpenAI data policy

Summarised from the provider's published terms. The provider's own policy is what applies.

Where you can run it

How GPT Transcribe can be deployed, bought and reached through Safeguard.

Deployment

Public cloudPrivate cloud

Procurement

Not published yet

Inference regions

Not published yet

Closed weights stay with the provider, so this model runs in the provider's cloud, never on-prem or air-gapped.

Pricing

Notes

Billed per audio minute: $0.0045 per minute. Not token-priced, so per-1M fields are null.

Source

OpenAI pricing , last verified October 8, 2026

No markup on model usage: you pay the provider's list price.

API

OpenAI-compatible. Your API key and access to this model are set up for your workspace when you choose Use with Safeguard.

from openai import OpenAI

client = OpenAI(
    base_url="https://api.safeguard.sh/gpt/v1",
    api_key="SAFEGUARD_API_KEY",
)

response = client.chat.completions.create(
    model="openai/gpt-transcribe",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

Base URL https://api.safeguard.sh/gpt/v1, model openai/gpt-transcribe.