GPT Audio
The gpt-audio model is OpenAI's first generally available audio model. The new snapshot features an upgraded decoder for more natural sounding voices and maintains better voice consistency.
The figures
- Made by
- OpenAI
- Model ID
- openai/gpt-audio
- Input
- $2.5 per million tokens
- Output
- $10 per million tokens
- Context window
- 128K tokens
- Longest answer
- 16K tokens
- Takes in
- text, audio
- Gives back
- text, audio
- Reasoning
- No
- Tool calling
- Yes
- Reads images
- No
- Open weights
- No
- Released
- 19 Jan 2026
Read on 3 Oct 2026.