click any column header to sort

TTS Bench — Speed — 2026-05-30_2330

Rig: windows-5090 — AMD Ryzen 9 9950X3D 16-Core Processor (16C) · NVIDIA GeForce RTX 5090 32GB · 126 GB RAM · Windows 11

Label: leaderboard-parity new models (cloning) — for merge into windows-cloning — 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): StyleTTS 2 (cuda) — 41.66× warm RTFx, 212ms warm TTFA

Model	Device	TTFA cold	TTFA warm	RTFx cold	RTFx warm	Peak RAM	Peak VRAM	Size	Released
StyleTTS 2	cuda	1.66s	212ms	5.45×	41.66×	2.63 GB	1.52 GB	~148M	Jun 2023
StyleTTS 2	cpu	1.70s	218ms	5.27×	40.62×	2.63 GB	—	~148M	Jun 2023
OpenVoice v2	cuda	1.51s	318ms	4.12×	18.95×	—	1.35 GB	~100M	Apr 2024
OpenVoice v2	cpu	2.25s	1.27s	2.57×	4.85×	—	—	~100M	Apr 2024
VibeVoice 7B	cuda	5.22s	4.86s	1.41×	1.58×	2.04 GB	17.63 GB	7B	Sep 2025
Fish Speech 1.5	cuda	8.04s	8.21s	0.85×	0.95×	3.82 GB	1.76 GB	~500M	Nov 2024
Zonos v0.1	cuda	10.36s	9.85s	0.67×	0.73×	6.55 GB	4.45 GB	1.6B	Feb 2025
Fish Speech 1.5	cpu	37.45s	42.66s	0.18×	0.18×	4.43 GB	—	~500M	Nov 2024
Zonos v0.1	cpu	61.60s	60.84s	0.12×	0.12×	8.57 GB	—	1.6B	Feb 2025