FluidVoice-Labs/fluid-speech-mini-onnx

Fluid Speech Mini (ONNX)

An English speech-to-text model built for dictation on phones and laptops. It matches the accuracy of Moondream's Parakeet Ultra (a 1.26 GB download) on the Open ASR Leaderboard's public English test sets, and was trained to keep transcribing the right person when someone else is talking in the room.

Fluid Speech Mini is the speech model inside FluidVoice. The iPhone app is in TestFlight: for access, send us a message on X at @FluidVoiceApp.

This is the ONNX build, for Windows and Linux: 302 MB, about 500 MB of memory on a 15-second clip, on any CPU or an NVIDIA GPU, with no PyTorch.

This is a port. Fluid Speech Mini was designed and trained for Apple devices, where Core ML stores its weights at about 3 bits each in 253 MB. ONNX Runtime has no 3-bit format, so this port stores the same layers at 2 and 4 bits. That is why the file is somewhat larger; accuracy is unchanged.

All builds

Build	For	Download	Memory, 15 s clip	Repository
Core ML	iPhone, Mac	253 MB	about 277 MB	FluidVoice-Labs/fluid-speech-mini-coreml
ONNX	Windows, Linux, CPU or NVIDIA GPU	302 MB	about 500 MB	FluidVoice-Labs/fluid-speech-mini-onnx
GGUF	Windows, Linux, C++ apps	347 MB	about 485 MB	FluidVoice-Labs/fluid-speech-mini-gguf

Numbers

Word error rate (WER) in percent; lower is better. All rows are our own runs with the Open ASR Leaderboard's evaluation code on identical clips, not entries on the public board.

Model	Download	Average, 8 public English sets	Another voice in the room
Fluid Speech Mini (Core ML build; this build: 4.53 and 11.1)	253 MB	4.54	10.9
Parakeet Ultra (Moondream)	1.26 GB	4.56	14.3
Parakeet TDT 0.6B v3 (NVIDIA)	2.51 GB	4.86	20.4
Phonon-2 (Fermion)	164 MB	5.13	24.6
Parakeet Redux (Moondream)	178 MB	5.62	25.7
Whisper large-v3	about 3 GB	5.77	11.8

Set by set: AMI 8.34, Earnings-22 6.29, GigaSpeech 7.67, LibriSpeech clean 1.45, LibriSpeech other 3.25, SPGISpeech 3.56, Monsoon 3.96, VoxPopuli 1.76.

"Another voice in the room" is Google SVQ (English, 28,004 phone recordings) in its background-speech condition. On the same clips with traffic noise, the larger Parakeet models and Whisper large-v3 score a couple of points better than this model; its advantage is other voices.

This build

This build holds weights fine-tuned from the Core ML model so that every packed row stores exactly in this format. It scores 4.53 % on the same eight test sets (Core ML: 4.54 %) and 11.1 % with another voice in the room (Core ML: 10.9 %). It gives the same transcript as the reference PyTorch model on all 5,559 utterances of LibriSpeech test-clean and test-other.

The weights stay packed in memory (ONNX Runtime MatMulNBits). Tested with ONNX Runtime 1.31 on Linux, Windows and macOS.

Use it

pip install onnxruntime numpy soundfile
hf download FluidVoice-Labs/fluid-speech-mini-onnx --local-dir fluid-speech-mini-onnx
python fluid-speech-mini-onnx/transcribe.py fluid-speech-mini-onnx recording.wav --threads 4

No PyTorch is needed. For an NVIDIA GPU, install onnxruntime-gpu and add --provider cuda. Audio that is not 16 kHz also needs soxr or scipy.

Limitations

• English only. The vocabulary is Parakeet v3's 25-language one, so other languages may come out, but they were not trained or measured and are expected to be worse.
• Its advantage in noise is specifically other voices; with traffic noise, larger models do better.
• Scores above are self-reported runs of the leaderboard's code, not entries on the public board.

License and attribution

CC BY 4.0. If you use the model, please credit it as Fluid Speech Mini by FluidVoice.

Built on NVIDIA Parakeet TDT 0.6B v3 and Moondream Parakeet Ultra, both CC BY 4.0. The decoder and joint network are Parakeet Ultra's.