cuDNN Frontend

by NVIDIA

C++ / Python API for cuDNN — speeds up custom ASR kernels.

TL;DR

C++ / Python API for cuDNN — speeds up custom ASR kernels.

Best for fused attention + convolution ops in Conformer/Whisper engines. Pricing: free.

Category
Open source
License
—
Stars
—
Last push
—
Pricing
free
Platforms
Linux

What it is

A modern wrapper around the cuDNN library. MIT.

Best for: Fused attention + convolution ops in Conformer/Whisper engines.
Watch out for: Low-level; expect to read examples.

Install / use

git clone https://github.com/NVIDIA/cudnn-frontend

Features

Speaker diarizationNo
Word-level timestampsYes
Streaming / real-timeNo
Languages supported99
HIPAA eligibleNo

cuDNN Frontend vs Whipscribe

FeaturecuDNN FrontendWhipscribe
CategoryOpen sourceTranscription APIs
Pricingfreefree beta
Speaker diarizationNoYes
Word timestampsYesYes
StreamingNoNo
Languages9999
PlatformsLinuxWeb, API, MCP

Alternatives to cuDNN Frontend

OpenAI Whisper
OpenAI

The reference open-source multilingual ASR model from OpenAI.

OSS · MIT ★ 98.1k
whisper.cpp
Georgi Gerganov

C/C++ port of Whisper — runs on anything, from a Raspberry Pi to Apple Silicon.

OSS · MIT ★ 48.8k
faster-whisper
SYSTRAN

4× faster than reference Whisper using CTranslate2 — production sweet spot.

OSS · MIT ★ 22.3k

Whipscribe is a managed faster-whisper + whisperX service. If you want transcripts without running infrastructure, paste a URL or drop a file in the form below — you'll have a transcript in seconds.