click any column header to sort

TTS Bench — Speed — 2026-05-25_0102

Rig: windows-5090 — AMD Ryzen 9 9950X3D 16-Core Processor (16C) · NVIDIA GeForce RTX 5090 32GB · 126 GB RAM · Windows 11
Label: patch · cloning · fixed + MOSS — 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): F5-TTS v1 (cuda) — 5.50× warm RTFx, 1.06s warm TTFA

ModelDeviceTTFA coldTTFA warmRTFx coldRTFx warmPeak RAMPeak VRAMSizeReleased
F5-TTS v1cuda1.56s1.06s3.75×5.50×2.67 GB883 MB330MOct 2024
MOSS-TTS-Nanocuda5.59s4.87s1.34×1.66×2.42 GB910 MB100MApr 2026
IndexTTS-2cpu5.77s4.87s1.20×1.47×5.54 GB1.5BJun 2025
MOSS-TTS v1.0cuda5.44s4.73s1.31×1.44×2.10 GB22.83 GB8BFeb 2026
IndexTTS-2cuda6.80s5.66s1.03×1.26×4.05 GB7.71 GB1.5BJun 2025
MOSS-TTS-Nanocpu5.69s5.04s1.16×1.25×3.05 GB100MApr 2026
Dia 1.6B-0626cuda13.79s11.45s0.37×0.50×4.42 GB4.82 GB1.6BJun 2025
F5-TTS v1cpu71.35s71.18s0.08×0.08×2.71 GB330MOct 2024