OpenAI · Transcription

GPT Transcribe

Supports context, keyword hints and multiple language hints

Provider
OpenAI
Family
GPT Audio
Model type
Transcription
Official release date
Jul 28, 2026
Release type
New model
Status
Generally available
Previous version
gpt-4o-transcribe / whisper-1
Input
audio, text
Output
text
Context window
Not documented
Max output
Not documented
API model id
gpt-transcribe
API price (input)
Not verified
API price (output)
Not verified
Price notes
$0.0045 per minute of audio (not per-token).
Consumer apps
—
Developer API
OpenAI API (audio/transcriptions, realtime transcription sessions)
Free access
None documented
Paid access
—
Availability
—
Licence
proprietary

Main improvements

  • Supports context, keyword hints and multiple language hints
  • Works for files and Realtime committed turns
  • Launched with GPT Live Transcribe for low-latency streaming

Best use cases

  • Transcribing recordings
  • Voice agents

Benchmarks

No benchmark results recorded. We only list scores with a source, benchmark version and date.

Verified Oct 19, 2026 · confidence: high

More from OpenAI

Compare with other models