Narrivy Store

Everything the Narrivy dictation app can add from its Store, in one place. Narrivy installs a pack only at a pinned revision of this repository, and checks every file against the pack's manifest before using it. Each pack's folder has its own README and licence.

Folder What is in it
models/gguf/ Speech models for the built-in engine (transcribe.cpp), on the CPU or a Vulkan GPU
models/openvino/ Speech models for the OpenVINO add-on, on an Intel NPU, GPU or the CPU
runtimes/openvino/ The OpenVINO add-on itself: the libraries and Intel's NPU driver
modules/ Add-on modules such as noise suppression and the writing assistant, as they are published

Packs

Pack Name Licence Size
models/gguf/parakeet-unified-en-0.6b-q8-0/ Parakeet Unified EN 0.6B (GGUF, Q8_0, 8-bit, streaming) CC-BY-4.0 731 MB
models/gguf/whisper-large-v3-q5-k-m/ Whisper large-v3 (GGUF, Q5_K_M, 5-bit) Apache-2.0 1,161 MB
models/gguf/whisper-large-v3-turbo-q8-0/ Whisper large-v3 turbo (GGUF, Q8_0, 8-bit) Apache-2.0 886 MB
models/openvino/whisper-large-v3-int8/ Whisper large-v3 (OpenVINO IR, INT8, 8-bit) Apache-2.0 1,568 MB
models/openvino/whisper-large-v3-turbo-int8/ Whisper large-v3 turbo (OpenVINO IR, INT8, 8-bit) MIT 828 MB
runtimes/openvino/2026.4-linux-x86_64/ OpenVINO runtime 2026.4.0 (linux-x86_64) Apache-2.0 AND MIT 292 MB
Downloads last month
11
GGUF
Model size
0.6B params
Architecture
parakeet
Hardware compatibility
Log In to add your hardware

5-bit

8-bit

Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support