Skip to content
Home›AI & agents›AI tools›Whisper (OpenAI)
Transcribe & subtitle

OpenAI Whisper: the open speech-to-text model other tools build on

OpenAI's open speech recognition model, MIT-licensed for code and weights, writing TXT, VTT, SRT, TSV and JSON [1][2].

Does
Transcripts, subtitles and translation into English
Best for
Tutorial and interview creators with a recent GPU, or time to wait on CPU
Cost
Free (MIT licence for code and weights) [1]
Runs
Locally; GPU recommended for larger models
Open the repositoryOfficial siteQuick start
Updated 2 Nov 2026

Whisper transcribes many languages, translates into English and identifies the language [1]. whisper.cpp, faster-whisper, WhisperX and Buzz build on it [5][6][7][8]. The whisper command works through each file in a sliding 30-second window [1].

What it does

Licence
MIT for code and model weights [1][3]
Output formats
txt, vtt, srt, tsv, json, jsonl, or all (the default), via --output_format [2]
Models
tiny (about 1 GB VRAM) to large (about 10 GB); turbo needs about 6 GB and runs about 8x faster than large [1]
Requirements
Python (3.8 to 3.11 expected to work), PyTorch, ffmpeg [1]
Translation
Into English with --task translate; turbo is not trained for translation [1]
Homebrew
openai-whisper formula, stable 20250625 [4]

Integrations

InterfaceStatusDetail
MCP serverMCP · CommunityCommunity wrappers
Command lineCLI · Built inPython CLI options
API / libraryAPI · Built inPython package

Requirements and setup

  • ffmpeg, such as brew install ffmpeg (macOS) or choco install ffmpeg (Windows) [1].
  • Rust, if tiktoken has no pre-built wheel for your platform [1].

Licence, cost and privacy

  • The package runs locally. OpenAI's hosted Whisper API is a separate remote service; Buzz can use it with an API key [9].

Limitations

  • Timestamps are per utterance, not per word, and can be several seconds out (WhisperX README) [7].
  • Slower on CPU than the ports: 13 minutes of audio, small model, 6m58s against 2m05s for whisper.cpp (faster-whisper's benchmark) [6].
  • Performance varies widely by language [1].
  • turbo returns the original language even with --task translate [1].
Quick start

Quick start

Install the package and ffmpeg, then run whisper on a file. turbo is the default model [1].

Copied from the documentation · source [1]
pip install -U openai-whisper
Copied from the documentation · macOS · source [4]
brew install openai-whisper
Copied from the documentation · source [1]
whisper audio.flac audio.mp3 audio.wav --model turbo
Assembled from documented options, not printed in the docs · source [2]
whisper audio.mp3 --model turbo --output_format srt
Copied from the documentation · source [1]
whisper japanese.wav --model medium --language Japanese --task translate

Without --output_format it writes every format [2]. whisper --help lists all options [1].

Uses for vloggers

  • Subtitles in the spoken language.
  • English subtitles for foreign-language footage, with a multilingual model [1].
  • A reference transcript for comparing the faster ports.

Alternatives

Transcribe & subtitlewhisper.cpp

Run efficient Whisper inference on local hardware

Free codeOn your computerDeveloperMIT licence
MCP · CommunityCLI · Built inAPI · Built in
Transcribe & subtitlefaster-whisper

Faster Whisper-based transcription with CTranslate2

Free codeOn your computerDeveloperMIT licence
MCP · CommunityCLI · CommunityAPI · Built in
Transcribe & subtitleWhisperX

Generate aligned word-level subtitles and diarisation workflows

Free codeOn your computerDeveloperBSD-2-Clause licence
MCP · CommunityCLI · Built inAPI · Built in
Transcribe & subtitleBuzz transcription

Desktop transcription / speech-to-text workflow

Free codeDesktop app · On your computerNo codingMIT licence
MCP · None foundCLI · Built inAPI · None found

Sources

  1. raw.githubusercontent.com/openai/whisper/main/README.md fetched 2 Nov 2026
  2. raw.githubusercontent.com/openai/whisper/main/whisper/transcribe.py fetched 2 Nov 2026
  3. raw.githubusercontent.com/openai/whisper/main/LICENSE fetched 2 Nov 2026
  4. Homebrew Formulae: openai-whisper fetched 2 Nov 2026
  5. raw.githubusercontent.com/ggml-org/whisper.cpp/master/README.md fetched 2 Nov 2026
  6. raw.githubusercontent.com/SYSTRAN/faster-whisper/master/README.md fetched 2 Nov 2026
  7. raw.githubusercontent.com/m-bain/whisperX/main/README.md fetched 2 Nov 2026
  8. raw.githubusercontent.com/chidiwilliams/buzz/main/README.md fetched 2 Nov 2026
  9. FAQ | Buzz fetched 2 Nov 2026