Datasets:
audio audioduration (s) 3.04 4.48 | request int64 1 10 | text stringclasses 5
values | words int64 5 9 | adapter stringclasses 5
values | concurrent_requests int64 200 200 | ttfa_ms float64 95.5 96.8 | burst_ttfa_p95_ms float64 96.8 96.8 | audio_s float64 3.04 4.48 | all_audio_ready_s float64 1.94 2.39 | chunks int64 21 30 | stall_s float64 0.15 0.15 | whisper_large_v3 stringclasses 5
values | cer_pct float64 0 23.7 | wer_pct float64 0 42.9 | speaker_similarity float64 0.85 0.93 | utmos float64 2.89 3.8 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1 | المكيف في الغرفة ما يبرد زين. | 6 | egyptian | 200 | 95.5 | 96.8 | 3.04 | 1.94 | 21 | 0.147 | المكيف في الغرفة ما يبرد زين | 0 | 0 | 0.909 | 3.08 | |
5 | كيفك؟ شو عم تعمل هالأيام؟ صرلنا زمان ما حكينا. | 9 | iraqi | 200 | 95.9 | 96.8 | 4.48 | 2.39 | 30 | 0.148 | كيفك؟ شو عم تعمل هالأيام؟ صرلنا زمان ما حكينا | 0 | 0 | 0.906 | 3 | |
7 | السيارة تبي لها بنزين، خلنا نوقف عند أقرب محطة. | 9 | najdi | 200 | 96.3 | 96.8 | 4 | 2.28 | 27 | 0.148 | السيارة تبي لها بنزين خلنا نوقف عند أقرب محطة | 0 | 0 | 0.927 | 2.89 | |
8 | ولدي بدأ يتعلم السباحة هالصيف. | 5 | palestinian | 200 | 96.4 | 96.8 | 3.52 | 2.13 | 24 | 0.149 | ولدي بدأ يتعلم السباحة الصيف | 3.4 | 20 | 0.852 | 3.8 | |
10 | أبي أحجز طاولة لأربعة أشخاص الساعة تسع. | 7 | saudi | 200 | 96.8 | 96.8 | 3.52 | 2.13 | 24 | 0.149 | أريد أن أحجز طاولة لأربعة أشخاص الساعة الساعة | 23.7 | 42.9 | 0.901 | 2.98 |
VoxCPM2 Arabic Dialects: time-to-first-audio samples
Five clips from a latency test: 200 voice-cloning requests sent at the same instant to one NVIDIA A10 GPU. Listen to
hear what the engine produces under that load; data/metadata.csv has every number per clip.
At a glance
| What | 5 generated clips and the reference voice they clone |
| Load | 200 requests at the same instant, one NVIDIA A10 (24 GB) |
| Time to first audio | 95.5-96.8 ms for these 5 clips; 96.8 ms p95 over all 200 requests |
| Dialect adapters | egyptian, iraqi, najdi, palestinian, saudi |
| Model | VoxCPM-Kernel/VoxCPM2-Arabic-Dialects (VoxCPM2 + 11 dialect LoRA adapters) |
| Engine | Speech-Next/voxcpm2-arabic-dialects-turbo |
| Measured | 2026-10-11 |
| License | CC BY 4.0 |
Time to first audio (TTFA) is the time from sending a request to receiving its first audio chunk, which is 40 ms of audio. The rest of the sentence streams in faster than it plays: a 3-4.5 s sentence is fully generated about 2 s after the request.
Samples
| file | sentence | words | TTFA | audio | whole sentence ready | CER | WER |
|---|---|---|---|---|---|---|---|
001_egyptian.wav |
المكيف في الغرفة ما يبرد زين. | 6 | 95.5 ms | 3.04 s | 1.94 s | 0.0 % | 0.0 % |
005_iraqi.wav |
كيفك؟ شو عم تعمل هالأيام؟ صرلنا زمان ما حكينا. | 9 | 95.9 ms | 4.48 s | 2.39 s | 0.0 % | 0.0 % |
007_najdi.wav |
السيارة تبي لها بنزين، خلنا نوقف عند أقرب محطة. | 9 | 96.3 ms | 4.00 s | 2.28 s | 0.0 % | 0.0 % |
008_palestinian.wav |
ولدي بدأ يتعلم السباحة هالصيف. | 5 | 96.4 ms | 3.52 s | 2.13 s | 3.4 % | 20.0 % |
010_saudi.wav |
أبي أحجز طاولة لأربعة أشخاص الساعة تسع. | 7 | 96.8 ms | 3.52 s | 2.13 s | 23.7 % | 42.9 % |
CER / WER: Whisper large-v3 transcript against the sentence, after Arabic normalization. All 200 outputs of the burst: CER 6.0 %, WER 14.7 %, UTMOS 2.95, speaker similarity 0.92.
Files
| path | content |
|---|---|
data/*.wav |
the 5 generated clips |
data/metadata.csv |
one row per clip: file_name, request, text, words, adapter, concurrent_requests, ttfa_ms, burst_ttfa_p95_ms, audio_s, all_audio_ready_s (request to last chunk), chunks, stall_s (playback wait after start), whisper_large_v3, cer_pct, wer_pct, speaker_similarity (WavLM-SV), utmos (1-5) |
reference/ar00.wav |
the voice every request clones |
How it was measured
| Engine settings | int8 W8A8 (SmoothQuant), 2-step distilled LocDiT, CUDA graphs; the first 40 ms frame is streamed ahead of the rest |
| Voice cloning | from the reference audio only, without its transcript |
| Requests | request i uses adapter i mod 11 and one of 192 short dialect sentences, so the adapter does not always match the sentence (the iraqi request reads a Levantine sentence) |
| Bursts | 1 warm-up, then 3 measured: p95 TTFA 96.8, 95.4 and 94.6 ms; the clips are from the first |
| Selection | the first request of each of the 5 adapters, not chosen by quality |
Load
from datasets import load_dataset
ds = load_dataset("VoxCPM-Kernel/VoxCPM2-Arabic-Dialects-Latency-Samples", split="train")
print(ds[0]["text"], ds[0]["ttfa_ms"])
License
CC BY 4.0. The reference voice reference/ar00.wav is a google/fleurs
clip (ar_eg train, id 1480, CC BY 4.0). The audio was generated with VoxCPM2 (Apache-2.0).
- Downloads last month
- 77