Vibeleaderboard
Index / tool
Visit github.com
Category
AI Tools
Rank
No. 3940Tools index

Previous survey · No. 3789 ·

Pricing
Open Source
Type
TOOL
Use case
Models: Train & Run
GitHub
5 stars
Date

About

Parakeetogo is a statically compiled speech transcription model packaged as a single binary with the model embedded, built for CPU-only deployment. It adapts Moondream's parakeet-redux model and uses Highway for portable SIMD acceleration. The author reports it runs faster than the Photon/torch version on CPU with nearly identical transcription quality.

At a glance

Parakeetogo runs a Parakeet-derived speech recognition model on a CPU through a standalone C++ engine. It is worth testing if you need local transcription without deploying a Python inference stack.

Good fit

Developers prototyping speech input on x86 or ARM hardware.

Consider something else when

Keep a supported transcription service if you need managed scaling or established accuracy for your languages and audio conditions.

Start here

Choose a few synthetic clips with different speeds and background noise. Check transcript errors and processing time on the target CPU.

Cost and requirements

Apache 2.0 for the project. Source builds use CMake. Check the model artifact and its applicable terms separately.

Limits to know

The project describes a quantized model and CPU-specific optimization. Measure accuracy and speed on your audio and hardware rather than relying on sample results.

Tags

asrtranscriptioncpusimdspeech-to-textc++

Tech Stack

Python

Comments (0)

No comments yet

Editorially curated, with community endorsements as a secondary signal. Corrections welcome.