click any column header to sort

TTS Bench — Speed — 2026-06-01_1003

Rig: windows-5090 — AMD Ryzen 9 9950X3D 16-Core Processor (16C) · NVIDIA GeForce RTX 5090 32GB · 126 GB RAM · Windows 10
Label: cloning · chris_hemsworth_15s · neutts_air,neutts_nano — ref chris_hemsworth_15s.wav
TTFA = time to first audio (ms; lower is better). RTFx = realtime speed (× realtime; higher is better; e.g. 10× means 10 sec of audio generated per 1 sec of compute). Cold = first run after process start; warm = subsequent runs.

Speed winners

Fastest (this cloning run): NeuTTS Nano (cuda) — 2.00× warm RTFx, 331ms warm TTFA

ModelDeviceTTFA coldTTFA warmRTFx coldRTFx warmPeak RAMPeak VRAMSizeReleased
NeuTTS Nanocuda854ms331ms1.60×2.00×2.59 GB3.26 GB229MDec 2025
NeuTTS Nanocpu863ms387ms1.23×1.58×5.12 GB229MDec 2025
NeuTTS Aircuda1.46s492ms0.88×1.30×2.96 GB3.26 GB748MSep 2025
NeuTTS Aircpu1.75s568ms0.75×0.96×5.47 GB748MSep 2025