Fact · Speech recognition
3.9GB of memory
The large model in whisper.cpp needs about 3.9 GB of memory and takes 2.9 GiB on disk.
Checked against its source on October 17, 2026
In the source's own words
“| large | 2.9 GiB | ~3.9 GB |”
whisper.cpp README, Georgi Gerganov, GitHub, 2026, Memory usage.
Open the original- What it measures
- The whisper.cpp README gives a memory table for each model. The row for tiny is about 273 MB of memory, and the row for medium is about 2.1 GB.
- What it does not show
- The table is the project's estimate. Actual memory use grows with audio length and the decoding settings.
Source
Cited on
- How to Run Whisper Locally on Your Own ComputerGuide
- AI Dictation Cost: Cloud vs On-Device EconomicsGuide
- Can AI dictation run entirely on your own computer?Question
- Can you run speech recognition on a Raspberry Pi?Question
- Does on-device AI dictation drain your battery?Question
- Is on-device dictation cheaper than cloud dictation?Question
- Is Whisper AI free?Question