Fact · Dictation in your apps
10GB of VRAM
The Whisper README gives about 10 GB of required VRAM for the large model, which the table rates as the 1x speed baseline.
Checked against its source on October 17, 2026
In the source's own words
“medium 769 M medium.en medium ~5 GB ~2x large 1550 M N/A large ~10 GB 1x”
openai/whisper: Robust Speech Recognition via Large-Scale Weak Supervision, OpenAI, GitHub (OpenAI repository), 2023, Available models and languages table.
Open the original- What it measures
- The large model has no English-only variant in the table (N/A), so it is the multilingual model only.
- What it does not show
- Approximate table value. A GPU with less memory cannot hold this model at full size, so tools fall back to smaller models or to the CPU.
Source
Cited on
- Is Whisper AI Safe? Privacy and Security ExplainedGuide
- Open-Source Speech Recognition: The Best ModelsGuide
- Speech to Text on Linux: Dictation on Ubuntu and MintGuide
- Transcribe Audio With Whisper, ChatGPT or GeminiGuide
- How do you use Whisper to transcribe audio?Question
- What is the best speech to text for Japanese?Question