Vendor page
turbo model release (openai/whisper discussion #2363)
OpenAI whisper maintainers, GitHub, openai/whisper repository, 2024
Read the originalRead on October 19, 2026
- What it is
- The OpenAI maintainer's release post on the Whisper repository, dated October 1, 2024, announcing the Whisper large-v3-turbo model. It describes the model as an optimized version of large-v3 with 4 decoder layers.
- How much weight it carries
- Written by a maintainer in the official repository, so it is the project's own announcement. It gives no benchmark figures, so the speed and accuracy claims in other sources are not confirmed here.
Facts taken from this source
- OpenAI's Whisper maintainers announced the large-v3-turbo model on October 1, 2024.
“Oct 1, 2024 Maintainer We’re releasing a new Whisper model named large-v3-turbo” (Release post dated Oct 1, 2024)
- The Whisper maintainers say large-v3-turbo is an optimized version of large-v3 with only 4 decoder layers, down from 32 in the large series.
“It is an optimized version of Whisper large-v3 and has only 4 decoder layers” (Release post)