GitLab Transcend - Intelligent orchestration
Mini toolsWhisper
By OpenAI
Whisper is an open-source speech model family built for multilingual automatic speech recognition, speech translation into English, and spoken language identification. The repository describes it as a general-purpose multitask Transformer seq2seq model trained on a large and diverse audio dataset, using special task tokens so one model can handle several speech-processing tasks in a unified framework. The current repo lists six size tiers, from 39M-parameter tiny models up to 1.55B-parameter large, plus a turbo model optimized from large-v3 for faster transcription with only a small accuracy tradeoff.
Multimodal
Gen 3
Overview
Whisper is OpenAI’s open-source general-purpose speech recognition model for transcription, translation, and language identification. It is a multitask Transformer sequence-to-sequence system trained on large-scale diverse audio, with multilingual support and model sizes ranging from tiny to large plus a faster turbo variant.
About OpenAI
OpenAI is a technology company that specializes in artificial intelligence research and innovation.
Industry:
Artificial Intelligence
Company Size:
4500
Location:
San Francisco, California, US
Website:
openai.com
Tools using Whisper
-
Transcribe audio & video with Whisper. Export TXT/SRT/VTT. Auto-delete 24h.Open
Last updated: April 6, 2026
