OpenAI Transcribe Models by OpenAI
Last refreshed 2026-10-17. Next refresh: weekly.
Details
Capabilities
Links
WebsiteAbout
OpenAI's speech-to-text model family. Includes the legacy whisper-1 (per-minute billing) and the newer gpt-4o-transcribe / gpt-4o-mini-transcribe / gpt-4o-transcribe-diarize (per-token billing with improved accuracy, language detection, and speaker diarization). Used via /v1/audio/transcriptions.
Decision facts
- Best fit
- speech recognitionaudiovision and multimodal work
- Capability starting point
- GPT-4o Transcribe Diarize with 16K context and multimodal inputs
- Lowest tracked input
- Not tracked
- Closest related family
- GPT-5
Current Variants
Use-when guidance is based on each model's tracked capabilities, context window, release date, and replacement status.
Use when the workload needs speech recognition, 16K context, and multimodal inputs.
Use when the workload needs speech recognition, 16K context, and multimodal inputs.
Use when the workload needs speech recognition, 16K context, and multimodal inputs.
Use when the workload needs speech recognition, multimodal inputs, and audio.
| Model | Use when | Released | Signals | Status |
|---|---|---|---|---|
| GPT-4o Transcribe Diarize | Use when the workload needs speech recognition, 16K context, and multimodal inputs. | 2025-10 | speech recognition16K contextmultimodal inputs | Current |
| GPT-4o Mini Transcribe | Use when the workload needs speech recognition, 16K context, and multimodal inputs. | 2025-03 | speech recognition16K contextmultimodal inputs | Current |
| GPT-4o Transcribe | Use when the workload needs speech recognition, 16K context, and multimodal inputs. | 2025-03 | speech recognition16K contextmultimodal inputs | Current |
| Whisper | Use when the workload needs speech recognition, multimodal inputs, and audio. | 2023-03 | speech recognitionmultimodal inputsaudio | Current |
Release Timeline
3 release groupsSpecifications(4 models)
| Model | Released | Context | Multimodal |
|---|---|---|---|
| GPT-4o Transcribe Diarize | 2025-10 | 16K | Yes |
| GPT-4o Mini Transcribe | 2025-03 | 16K | Yes |
| GPT-4o Transcribe | 2025-03 | 16K | Yes |
| Whisper | 2023-03 | — | Yes |
