Research + corpus-building uses. Benchmarks don’t count.
The drawers
A tool can be filed in more than one drawer
DRAWER I
AI & History
Machine learning, vision-language models and speech models. Every card here carries notes on where your sources go, how it was tested, and how it fails.
Guide cards
Counts research and corpus-building uses in the window, plus “I’m using this now” check-ins. Benchmarks don’t count.
tool · No. 027AI
Whisper
Speech-recognition model that transcribes oral history interviews and archival audio, with timestamps, in many languages.
At a glance
Maker
OpenAI
Released
Sep 2022
Licence
Open source (MIT)
Cost
Free to run yourself; OpenAI API is paid
Runs
Your own machine (Python); OpenAI API
Skill
Some code
Historian’s notes
Where your sources goRun locally, audio never leaves your machine. The API sends it to OpenAI, where it may be kept up to 30 days.
How it’s been tested15.3% average word error across 1,847 Holocaust testimonies in seven languages (Fortunoff Archive, 2026).
Known problems on historical sourcesCan invent whole sentences nobody said, especially around pauses. Mishears dialect and proper names. Always check against the recording.
Used in research0
No published research uses found yet. Using it in a project? Tell us.
Nobody has said they’re using it in 2026 yet.
“I’m using this now” asks you to sign in with ORCID.
One count per person per year; it adds to Trending, and names are never shown.