# Whisper large-v3 requirements

Updated: 2026-06-15

1.55B Speech to text, released 2023-11.

- Peak memory: ~2.5 GB at int8
- Tools: whisper.cpp, faster-whisper, MacWhisper, WhisperX
- License: MIT (commercial use: yes)
- Runs on 39 of 39 tracked devices

Sources: https://github.com/openai/whisper, https://huggingface.co/openai/whisper-large-v3, https://github.com/SYSTRAN/faster-whisper, https://github.com/ggml-org/whisper.cpp
More: https://localmodel.run/model/whisper-large-v3