# Whisper

> machine learning model for speech recognition and transcription

**Wikidata**: [Q120759195](https://www.wikidata.org/wiki/Q120759195)  
**Wikipedia**: [English](https://en.wikipedia.org/wiki/Whisper_(speech_recognition_system))  
**Source**: https://4ort.xyz/entity/whisper

## Summary
Whisper is a machine learning model developed by OpenAI for speech recognition and transcription. It functions as a Python package and software component designed to convert spoken language into text. The model is accessible via GitHub and is distributed under the MIT License.

## Key Facts
- **Developer:** OpenAI and OpenAI OpCo.
- **License:** MIT License.
- **Category:** Instance of a Python package, software, and machine learning model.
- **Repository:** Hosted at `https://github.com/openai/whisper`.
- **PyPI Project:** Listed as `openai-whisper`.
- **Aliases:** Also known as OpenAI Whisper, WhisperAI, and openai-whisper.
- **Copyright Status:** Copyrighted.
- **Wikipedia Presence:** Available in 10 languages including English, Spanish, French, Japanese, and Arabic.

## FAQs
### Q: Who created Whisper?
A: Whisper was created and developed by OpenAI, an American artificial intelligence research organization, and its for-profit subsidiary, OpenAI OpCo.

### Q: What software license is Whisper released under?
A: Whisper is released under the MIT License, as confirmed by its GitHub repository.

### Q: How can developers access Whisper?
A: Developers can access the source code via GitHub or install it as a Python package through the PyPI project `openai-whisper`.

### Q: What type of technology is Whisper?
A: Whisper is a machine learning model classified as software and a Python package, specifically designed for speech recognition and transcription tasks.

## Why It Matters
Whisper represents a significant contribution to the field of automated speech recognition (ASR). Developed by OpenAI, a leading entity in the artificial intelligence industry, it provides a robust tool for transcribing audio into text. Unlike proprietary "black box" solutions, Whisper is available as a Python package with a permissive MIT License, allowing researchers and developers to inspect, modify, and integrate the model into various applications with few restrictions. This accessibility fosters innovation in voice-enabled technologies, ranging from accessibility tools to automated transcription services.

The model's development by a major AI research organization ensures it is built upon substantial machine learning infrastructure. By open-sourcing the model via GitHub, OpenAI has provided a valuable resource for the tech community, standardizing high-quality transcription capabilities. Its availability in multiple Wikipedia languages and its recognition as a distinct software entity underscore its global utility and adoption in the software ecosystem.

## Notable For
- **Open Source Accessibility:** Unlike many high-performance models kept behind API gateways, Whisper is available under the MIT License.
- **Origin:** It is a product of OpenAI, a prominent organization in the AI sector known for breakthroughs in generative AI.
- **Integration:** It is distributed as a standard Python package (`openai-whisper`), facilitating easy integration into existing software workflows.
- **Functionality:** Specifically designed for the complex tasks of speech recognition and transcription.
- **Global Reach:** Documentation and information regarding the system are available across a diverse set of languages, including Arabic, Persian, and Ukrainian.

## Body

### Development and Background
Whisper is a machine learning model developed by **OpenAI** and **OpenAI OpCo**. OpenAI is an American artificial intelligence research organization founded in December 2015, while OpenAI OpCo serves as its for-profit subsidiary, established in September 2018. Both entities are headquartered in San Francisco, California, within the Pioneer Building.

The development of Whisper aligns with OpenAI's industry focus on artificial intelligence. The organization has grown significantly, reporting 375 employees as of early 2023. Whisper operates as a tangible executable component of the computer software landscape, distinct from non-tangible research concepts.

### Technical Specifications
Whisper is classified as a **machine learning model**, **software**, and **Python package**. Its primary function is speech recognition and transcription. The system is identified by several aliases, including "OpenAI Whisper," "WhisperAI," and "openai-whisper."

*   **Source Code:** The code is maintained in a repository hosted at `github.com/openai/whisper`.
*   **Distribution:** It is available for installation via the Python Package Index (PyPI) under the project name `openai-whisper`.
*   **Copyright:** The software is copyrighted, though it is distributed freely under the MIT License.

### Licensing and Usage
The use of Whisper is governed by the **MIT License**, a permissive free software license originating from the Massachusetts Institute of Technology. This licensing allows for wide reuse and adaptation of the software. The license status was verified via the GitHub API in July 2023.

### Digital Presence
Whisper maintains a significant digital footprint as evidenced by its presence on multiple platforms. It has a dedicated Wikipedia entry titled "Whisper (speech recognition system)" which is available in 10 distinct languages, including English, French, Japanese, Portuguese, and Catalan. It is also tracked on AlternativeTo software discovery services.

## References

1. [Source](https://api.github.com/repos/openai/whisper)