click any column header to sort
← SpeedSpeedSamples → · all runs

TTS Bench — Samples — windows-cloning

Rig: windows-5090 — AMD Ryzen 9 9950X3D 16-Core Processor (16C) · NVIDIA GeForce RTX 5090 32GB · 126 GB RAM · Windows 11
Label: cloning · chris_hemsworth_15s — ref chris_hemsworth_15s.wav
5 prompt(s) · one section per prompt · all models ranked by warm TTFA (fastest first) within each
Each prompt section shows every model's audio output, ordered by warm TTFA (fastest first). Click any audio player to hear that model's rendering.

Reference voice

Each model below was given this clip + transcript as the voice to imitate. Source: chris_hemsworth_15s.wav

Prompt 1

[en]"Open the browser and read my email."
Rank Model Device TTFA warm Audio
1 Pocket-TTS cpu 63ms
2 StyleTTS 2 cpu 129ms
3 StyleTTS 2 cuda 140ms
4 OpenVoice v2 cuda 185ms
5 NeuTTS Nano cuda 356ms
6 NeuTTS Nano cpu 369ms
7 NeuTTS Air cuda 497ms
8 Qwen3-TTS 1.7B (CUDA-graph) cuda 529ms
9 Coqui XTTS-v2 cuda 563ms
10 OpenVoice v2 cpu 623ms
11 NeuTTS Air cpu 630ms
12 Chatterbox Turbo cuda 632ms
13 MiraTTS cuda 645ms
14 F5-TTS v1 cuda 728ms
15 OmniVoice cuda 736ms
16 ZipVoice 123M cuda 1.08s
17 Chatterbox cuda 1.25s
18 Dia 1.6B-0626 cuda 1.30s
19 LongCat-AudioDiT 1B cuda 1.44s
20 VibeVoice 7B cuda 1.68s
21 VoxCPM2 2B cuda 1.81s
22 WavTTS 0.67B cuda 1.97s
23 IndexTTS-2 cpu 2.09s
24 Fish Speech 1.5 cuda 2.16s
25 LongCat-AudioDiT 3.5B cuda 2.19s
26 MOSS-TTS-Nano cuda 2.28s
27 Echo-TTS cuda 2.29s
28 Coqui XTTS-v2 cpu 2.53s
29 IndexTTS-2 cuda 2.57s
30 MOSS-TTS-Nano cpu 2.61s
31 MOSS-TTS v1.5 cuda 2.98s
32 MOSS-TTS v1.0 cuda 2.99s
33 Zonos v0.1 cuda 3.41s
34 Chatterbox Turbo cpu 3.48s
35 Sesame CSM-1B cuda 3.97s
36 DramaBox cuda 4.12s
37 OuteTTS 1.0 1B cuda 6.26s
38 VoxCPM2 2B cpu 6.39s
39 Miso TTS 8B cuda 6.46s
40 Chatterbox cpu 6.52s
41 ZipVoice 123M cpu 7.67s
42 Fish Speech 1.5 cpu 11.56s
43 Sesame CSM-1B cpu 17.17s
44 Zonos v0.1 cpu 18.60s
45 Mars5-TTS cuda 29.79s
46 Mars5-TTS cpu 30.56s
47 OmniVoice cpu 37.89s
48 F5-TTS v1 cpu 50.19s

Prompt 2

[en]"I'll start a new git branch, push the changes, and open a pull request when the tests pass."
Rank Model Device TTFA warm Audio
1 Pocket-TTS cpu 86ms
2 StyleTTS 2 cuda 173ms
3 StyleTTS 2 cpu 173ms
4 OpenVoice v2 cuda 339ms
5 NeuTTS Nano cuda 368ms
6 NeuTTS Nano cpu 460ms
7 NeuTTS Air cpu 495ms
8 NeuTTS Air cuda 496ms
9 OmniVoice cuda 759ms
10 F5-TTS v1 cuda 914ms
11 MiraTTS cuda 1.08s
12 Coqui XTTS-v2 cuda 1.09s
13 OpenVoice v2 cpu 1.09s
14 Chatterbox Turbo cuda 1.13s
15 Qwen3-TTS 1.7B (CUDA-graph) cuda 1.22s
16 LongCat-AudioDiT 1B cuda 1.28s
17 LongCat-AudioDiT 3.5B cuda 1.93s
18 Echo-TTS cuda 2.16s
19 WavTTS 0.67B cuda 2.19s
20 Chatterbox cuda 2.21s
21 MOSS-TTS-Nano cuda 3.38s
22 VoxCPM2 2B cuda 3.43s
23 VibeVoice 7B cuda 3.50s
24 MOSS-TTS v1.0 cuda 3.52s
25 MOSS-TTS v1.5 cuda 3.83s
26 MOSS-TTS-Nano cpu 3.97s
27 IndexTTS-2 cpu 4.01s
28 IndexTTS-2 cuda 4.12s
29 DramaBox cuda 4.63s
30 Fish Speech 1.5 cuda 5.49s
31 Coqui XTTS-v2 cpu 6.01s
32 Chatterbox Turbo cpu 6.28s
33 Zonos v0.1 cuda 7.85s
34 Dia 1.6B-0626 cuda 8.34s
35 Sesame CSM-1B cuda 9.01s
36 Chatterbox cpu 11.99s
37 ZipVoice 123M cpu 13.40s
38 OuteTTS 1.0 1B cuda 13.48s
39 VoxCPM2 2B cpu 14.44s
40 Fish Speech 1.5 cpu 30.34s
41 Miso TTS 8B cuda 33.31s
42 Mars5-TTS cuda 33.89s
43 Sesame CSM-1B cpu 38.15s
44 Mars5-TTS cpu 43.83s
45 Zonos v0.1 cpu 47.31s
46 OmniVoice cpu 48.65s
47 F5-TTS v1 cpu 64.77s
48 ZipVoice 123M cuda 133.84s

Prompt 3

[en]"The Parakeet TDT zero point six billion parameter model achieves one point six nine percent word error rate on LibriSpeech test-clean, beating Whisper Large V3 at two point seven percent while running at over two thousand times realtime on a single GPU."
Rank Model Device TTFA warm Audio
1 Pocket-TTS cpu 105ms
2 StyleTTS 2 cuda 306ms
3 NeuTTS Nano cuda 307ms
4 StyleTTS 2 cpu 340ms
5 NeuTTS Nano cpu 378ms
6 NeuTTS Air cuda 518ms
7 OpenVoice v2 cuda 525ms
8 NeuTTS Air cpu 636ms
9 OmniVoice cuda 934ms
10 LongCat-AudioDiT 1B cuda 1.38s
11 F5-TTS v1 cuda 1.71s
12 Echo-TTS cuda 2.14s
13 MiraTTS cuda 2.62s
14 LongCat-AudioDiT 3.5B cuda 2.81s
15 OpenVoice v2 cpu 3.05s
16 Chatterbox Turbo cuda 3.27s
17 Coqui XTTS-v2 cuda 3.35s
18 Qwen3-TTS 1.7B (CUDA-graph) cuda 3.37s
19 WavTTS 0.67B cuda 4.91s
20 Chatterbox cuda 5.20s
21 DramaBox cuda 6.22s
22 MOSS-TTS v1.0 cuda 8.96s
23 IndexTTS-2 cpu 9.05s
24 MOSS-TTS-Nano cuda 9.17s
25 MOSS-TTS v1.5 cuda 9.43s
26 VibeVoice 7B cuda 9.65s
27 VoxCPM2 2B cuda 10.29s
28 MOSS-TTS-Nano cpu 10.40s
29 IndexTTS-2 cuda 10.75s
30 Sesame CSM-1B cuda 17.44s
31 Chatterbox Turbo cpu 20.05s
32 Fish Speech 1.5 cuda 20.28s
33 Coqui XTTS-v2 cpu 20.33s
34 Zonos v0.1 cuda 23.91s
35 Dia 1.6B-0626 cuda 24.69s
36 Chatterbox cpu 34.74s
37 VoxCPM2 2B cpu 35.11s
38 OuteTTS 1.0 1B cuda 39.66s
39 Mars5-TTS cpu 54.81s
40 Mars5-TTS cuda 55.86s
41 Miso TTS 8B cuda 68.73s
42 Sesame CSM-1B cpu 77.75s
43 OmniVoice cpu 78.95s
44 F5-TTS v1 cpu 103.89s
45 Fish Speech 1.5 cpu 107.15s
46 Zonos v0.1 cpu 157.87s
47 ZipVoice 123M cpu 183.07s

Prompt 4

[en]"Run pytest tests slash test underscore voice dot py with verbose flag and capture flag set to no."
Rank Model Device TTFA warm Audio
1 Pocket-TTS cpu 95ms
2 StyleTTS 2 cuda 228ms
3 StyleTTS 2 cpu 230ms
4 NeuTTS Nano cuda 337ms
5 OpenVoice v2 cuda 378ms
6 NeuTTS Nano cpu 386ms
7 NeuTTS Air cuda 458ms
8 NeuTTS Air cpu 512ms
9 OmniVoice cuda 752ms
10 F5-TTS v1 cuda 870ms
11 OpenVoice v2 cpu 1.30s
12 Chatterbox Turbo cuda 1.32s
13 LongCat-AudioDiT 1B cuda 1.35s
14 MiraTTS cuda 1.48s
15 Qwen3-TTS 1.7B (CUDA-graph) cuda 1.54s
16 Coqui XTTS-v2 cuda 1.59s
17 WavTTS 0.67B cuda 2.29s
18 Echo-TTS cuda 2.34s
19 LongCat-AudioDiT 3.5B cuda 2.34s
20 Chatterbox cuda 2.87s
21 MOSS-TTS-Nano cuda 3.69s
22 IndexTTS-2 cpu 4.32s
23 MOSS-TTS-Nano cpu 4.58s
24 VibeVoice 7B cuda 4.60s
25 MOSS-TTS v1.0 cuda 4.69s
26 DramaBox cuda 4.98s
27 IndexTTS-2 cuda 5.19s
28 VoxCPM2 2B cuda 5.19s
29 MOSS-TTS v1.5 cuda 6.51s
30 Fish Speech 1.5 cuda 7.20s
31 Chatterbox Turbo cpu 7.65s
32 Zonos v0.1 cuda 8.48s
33 Coqui XTTS-v2 cpu 9.36s
34 Dia 1.6B-0626 cuda 10.53s
35 VoxCPM2 2B cpu 14.38s
36 Sesame CSM-1B cuda 15.42s
37 Chatterbox cpu 16.32s
38 ZipVoice 123M cpu 16.61s
39 OuteTTS 1.0 1B cuda 17.15s
40 Miso TTS 8B cuda 21.77s
41 Fish Speech 1.5 cpu 36.09s
42 Mars5-TTS cuda 38.59s
43 Mars5-TTS cpu 43.71s
44 Zonos v0.1 cpu 49.51s
45 OmniVoice cpu 49.71s
46 Sesame CSM-1B cpu 61.47s
47 F5-TTS v1 cpu 65.87s
48 ZipVoice 123M cuda 142.27s

Prompt 5

[fr]"Bonjour, je m'appelle Cicero et je vais vous aider avec votre code aujourd'hui."
Rank Model Device TTFA warm Audio
1 Pocket-TTS cpu 273ms
2 NeuTTS Nano cuda 289ms
3 NeuTTS Nano cpu 342ms
4 OpenVoice v2 cuda 379ms
5 OmniVoice cuda 833ms
6 Coqui XTTS-v2 cuda 953ms
7 OpenVoice v2 cpu 1.12s
8 Qwen3-TTS 1.7B (CUDA-graph) cuda 1.20s
9 VoxCPM2 2B cuda 2.87s
10 MOSS-TTS v1.0 cuda 3.52s
11 MOSS-TTS-Nano cpu 3.61s
12 MOSS-TTS v1.5 cuda 3.64s
13 Zonos v0.1 cuda 5.57s
14 Coqui XTTS-v2 cpu 5.58s
15 MOSS-TTS-Nano cuda 5.84s
16 Fish Speech 1.5 cuda 5.93s
17 VoxCPM2 2B cpu 9.90s
18 OuteTTS 1.0 1B cuda 10.67s
19 ZipVoice 123M cpu 14.56s
20 Fish Speech 1.5 cpu 28.17s
21 Zonos v0.1 cpu 30.92s
22 OmniVoice cpu 51.20s
23 ZipVoice 123M cuda 146.78s