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

TTS Bench — Samples — windows-default

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

Prompt 1

[en]"Open the browser and read my email."
Rank Model Device TTFA warm Audio
1 Kokoro cuda 47ms
2 Piper cpu 47ms
3 Pocket-TTS cpu 66ms
4 MeloTTS cuda 75ms
5 LFM2.5-Audio 1.5B cpu 83ms
6 LFM2.5-Audio 1.5B cuda 89ms
7 StyleTTS 2 cpu 168ms
8 StyleTTS 2 cuda 175ms
9 Kokoro cpu 226ms
10 NeuTTS Nano cuda 270ms
11 NeuTTS Nano cpu 303ms
12 OpenVoice v2 cuda 304ms
13 MeloTTS cpu 345ms
14 NeuTTS Air cuda 411ms
15 Supertonic 3 cpu 438ms
16 NeuTTS Air cpu 463ms
17 KittenTTS Nano 0.1 cpu 504ms
18 Qwen3-TTS 1.7B (CUDA-graph) cuda 557ms
19 Soprano 1.1 80M cuda 557ms
20 OpenVoice v2 cpu 595ms
21 Coqui XTTS-v2 cuda 614ms
22 Soprano 1.1 80M cpu 661ms
23 Chatterbox Turbo cuda 725ms
24 F5-TTS v1 cuda 778ms
25 OmniVoice cuda 887ms
26 ZipVoice 123M cuda 1.20s
27 Chatterbox cuda 1.22s
28 LongCat-AudioDiT 1B cuda 1.24s
29 VibeVoice Realtime 0.5B cuda 1.35s
30 Magpie-TTS cuda 1.61s
31 LongCat-AudioDiT 3.5B cuda 1.76s
32 VibeVoice 1.5B cuda 1.85s
33 IndexTTS-2 cpu 2.04s
34 Echo-TTS cuda 2.11s
35 VoxCPM2 2B cuda 2.48s
36 MOSS-TTS-Nano cuda 2.57s
37 Parler-TTS Mini v1 cuda 2.58s
38 MOSS-TTS v1.0 cuda 2.72s
39 IndexTTS-2 cuda 2.74s
40 Coqui XTTS-v2 cpu 2.75s
41 Qwen3-TTS 1.7B Base cuda 2.78s
42 MOSS-TTS-Nano cpu 2.99s
43 DramaBox cuda 2.99s
44 Fish Speech 1.5 cuda 3.06s
45 Zonos v0.1 cuda 3.50s
46 Chatterbox Turbo cpu 3.62s
47 Sesame CSM-1B cuda 3.73s
48 VibeVoice Realtime 0.5B cpu 4.74s
49 Maya1 cuda 4.75s
50 Dia 1.6B-0626 cuda 4.75s
51 Chatterbox cpu 5.32s
52 Miso TTS 8B cuda 5.48s
53 MOSS-TTS v1.5 cuda 5.62s
54 Magpie-TTS cpu 7.12s
55 OmniVoice cpu 7.50s
56 VibeVoice 7B cuda 7.64s
57 VoxCPM2 2B cpu 8.44s
58 OuteTTS 1.0 1B cuda 10.20s
59 ZipVoice 123M cpu 10.63s
60 Qwen3-TTS 1.7B Base cpu 11.69s
61 Fish Speech 1.5 cpu 15.07s
62 Parler-TTS Mini v1 cpu 16.48s
63 Mars5-TTS cpu 19.20s
64 Mars5-TTS cuda 19.75s
65 Zonos v0.1 cpu 19.96s
66 VibeVoice 1.5B cpu 20.22s
67 Sesame CSM-1B cpu 20.89s
68 F5-TTS v1 cpu 45.35s
69 Maya1 cpu 49.55s

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 Kokoro cuda 59ms
2 Piper cpu 89ms
3 LFM2.5-Audio 1.5B cpu 89ms
4 Pocket-TTS cpu 94ms
5 LFM2.5-Audio 1.5B cuda 103ms
6 MeloTTS cuda 107ms
7 StyleTTS 2 cuda 188ms
8 StyleTTS 2 cpu 198ms
9 NeuTTS Nano cuda 268ms
10 NeuTTS Nano cpu 317ms
11 OpenVoice v2 cuda 353ms
12 Kokoro cpu 380ms
13 NeuTTS Air cuda 420ms
14 NeuTTS Air cpu 506ms
15 Supertonic 3 cpu 586ms
16 MeloTTS cpu 656ms
17 F5-TTS v1 cuda 735ms
18 OmniVoice cuda 773ms
19 KittenTTS Nano 0.1 cpu 843ms
20 LongCat-AudioDiT 1B cuda 971ms
21 Coqui XTTS-v2 cuda 1.15s
22 OpenVoice v2 cpu 1.18s
23 Qwen3-TTS 1.7B (CUDA-graph) cuda 1.30s
24 Chatterbox Turbo cuda 1.30s
25 Soprano 1.1 80M cuda 1.33s
26 Soprano 1.1 80M cpu 1.43s
27 LongCat-AudioDiT 3.5B cuda 1.49s
28 Echo-TTS cuda 2.20s
29 Chatterbox cuda 2.24s
30 VibeVoice Realtime 0.5B cuda 2.64s
31 DramaBox cuda 3.42s
32 Magpie-TTS cuda 3.45s
33 VibeVoice 1.5B cuda 3.47s
34 VoxCPM2 2B cuda 3.59s
35 VibeVoice 7B cuda 3.78s
36 MOSS-TTS v1.0 cuda 3.94s
37 IndexTTS-2 cuda 4.16s
38 IndexTTS-2 cpu 4.23s
39 MOSS-TTS-Nano cpu 4.47s
40 MOSS-TTS-Nano cuda 5.23s
41 MOSS-TTS v1.5 cuda 5.89s
42 Parler-TTS Mini v1 cuda 6.04s
43 Qwen3-TTS 1.7B Base cuda 6.47s
44 Coqui XTTS-v2 cpu 6.65s
45 Fish Speech 1.5 cuda 7.07s
46 Zonos v0.1 cuda 7.78s
47 Chatterbox Turbo cpu 8.06s
48 VibeVoice Realtime 0.5B cpu 9.36s
49 Chatterbox cpu 10.32s
50 Sesame CSM-1B cuda 10.43s
51 Maya1 cuda 10.54s
52 VoxCPM2 2B cpu 11.56s
53 OmniVoice cpu 13.09s
54 OuteTTS 1.0 1B cuda 20.98s
55 Magpie-TTS cpu 21.02s
56 Miso TTS 8B cuda 22.52s
57 Qwen3-TTS 1.7B Base cpu 22.87s
58 Mars5-TTS cuda 25.09s
59 Mars5-TTS cpu 26.10s
60 VibeVoice 1.5B cpu 29.51s
61 Fish Speech 1.5 cpu 31.14s
62 Parler-TTS Mini v1 cpu 42.46s
63 Zonos v0.1 cpu 45.49s
64 Dia 1.6B-0626 cuda 51.96s
65 F5-TTS v1 cpu 58.82s
66 Sesame CSM-1B cpu 66.45s
67 Maya1 cpu 97.77s
68 ZipVoice 123M cuda 137.13s

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 LFM2.5-Audio 1.5B cuda 79ms
2 LFM2.5-Audio 1.5B cpu 84ms
3 Pocket-TTS cpu 109ms
4 Kokoro cuda 116ms
5 MeloTTS cuda 153ms
6 Piper cpu 239ms
7 NeuTTS Nano cuda 265ms
8 NeuTTS Nano cpu 322ms
9 StyleTTS 2 cpu 328ms
10 StyleTTS 2 cuda 374ms
11 NeuTTS Air cuda 423ms
12 NeuTTS Air cpu 458ms
13 OpenVoice v2 cuda 485ms
14 OmniVoice cuda 714ms
15 F5-TTS v1 cuda 1.03s
16 Kokoro cpu 1.27s
17 LongCat-AudioDiT 1B cuda 1.29s
18 Supertonic 3 cpu 1.45s
19 MeloTTS cpu 1.76s
20 Echo-TTS cuda 2.20s
21 LongCat-AudioDiT 3.5B cuda 2.20s
22 KittenTTS Nano 0.1 cpu 2.46s
23 OpenVoice v2 cpu 3.11s
24 Chatterbox Turbo cuda 3.12s
25 Qwen3-TTS 1.7B (CUDA-graph) cuda 3.30s
26 Soprano 1.1 80M cuda 3.70s
27 Soprano 1.1 80M cpu 4.27s
28 Chatterbox cuda 4.68s
29 Coqui XTTS-v2 cuda 4.82s
30 DramaBox cuda 5.19s
31 VibeVoice Realtime 0.5B cuda 7.86s
32 MOSS-TTS v1.0 cuda 8.40s
33 VibeVoice 7B cuda 8.77s
34 Magpie-TTS cuda 9.64s
35 MOSS-TTS-Nano cuda 10.45s
36 VibeVoice 1.5B cuda 10.63s
37 VoxCPM2 2B cuda 11.06s
38 IndexTTS-2 cpu 11.13s
39 MOSS-TTS-Nano cpu 11.83s
40 MOSS-TTS v1.5 cuda 11.88s
41 IndexTTS-2 cuda 12.58s
42 Fish Speech 1.5 cuda 17.39s
43 Parler-TTS Mini v1 cuda 18.07s
44 Sesame CSM-1B cuda 18.23s
45 Chatterbox Turbo cpu 19.26s
46 Qwen3-TTS 1.7B Base cuda 19.42s
47 Dia 1.6B-0626 cuda 22.29s
48 Coqui XTTS-v2 cpu 25.28s
49 Zonos v0.1 cuda 25.58s
50 VibeVoice Realtime 0.5B cpu 26.34s
51 Chatterbox cpu 28.27s
52 Maya1 cuda 29.31s
53 VoxCPM2 2B cpu 32.23s
54 OmniVoice cpu 34.51s
55 OuteTTS 1.0 1B cuda 50.39s
56 Mars5-TTS cpu 50.94s
57 Mars5-TTS cuda 51.19s
58 Qwen3-TTS 1.7B Base cpu 63.06s
59 Miso TTS 8B cuda 75.59s
60 F5-TTS v1 cpu 77.80s
61 Sesame CSM-1B cpu 78.74s
62 VibeVoice 1.5B cpu 85.74s
63 Magpie-TTS cpu 87.47s
64 Fish Speech 1.5 cpu 96.22s
65 Parler-TTS Mini v1 cpu 140.05s
66 Zonos v0.1 cpu 156.70s

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 Kokoro cuda 65ms
2 LFM2.5-Audio 1.5B cpu 78ms
3 LFM2.5-Audio 1.5B cuda 91ms
4 Pocket-TTS cpu 92ms
5 Piper cpu 93ms
6 MeloTTS cuda 108ms
7 NeuTTS Nano cuda 258ms
8 NeuTTS Nano cpu 297ms
9 StyleTTS 2 cpu 309ms
10 StyleTTS 2 cuda 323ms
11 OpenVoice v2 cuda 387ms
12 NeuTTS Air cuda 417ms
13 Kokoro cpu 452ms
14 NeuTTS Air cpu 456ms
15 Supertonic 3 cpu 663ms
16 OmniVoice cuda 694ms
17 MeloTTS cpu 743ms
18 F5-TTS v1 cuda 836ms
19 KittenTTS Nano 0.1 cpu 1.01s
20 LongCat-AudioDiT 1B cuda 1.26s
21 Chatterbox Turbo cuda 1.33s
22 OpenVoice v2 cpu 1.39s
23 Soprano 1.1 80M cuda 1.49s
24 Soprano 1.1 80M cpu 1.62s
25 Qwen3-TTS 1.7B (CUDA-graph) cuda 1.66s
26 LongCat-AudioDiT 3.5B cuda 1.74s
27 Coqui XTTS-v2 cuda 1.93s
28 Echo-TTS cuda 2.09s
29 Chatterbox cuda 2.31s
30 VibeVoice Realtime 0.5B cuda 3.23s
31 DramaBox cuda 3.52s
32 MOSS-TTS-Nano cuda 3.67s
33 IndexTTS-2 cpu 4.29s
34 Magpie-TTS cuda 4.37s
35 VibeVoice 7B cuda 4.48s
36 IndexTTS-2 cuda 4.64s
37 MOSS-TTS v1.0 cuda 4.90s
38 MOSS-TTS-Nano cpu 5.00s
39 VibeVoice 1.5B cuda 5.30s
40 Parler-TTS Mini v1 cuda 5.87s
41 VoxCPM2 2B cuda 6.05s
42 Fish Speech 1.5 cuda 6.90s
43 MOSS-TTS v1.5 cuda 7.04s
44 Chatterbox Turbo cpu 8.76s
45 Zonos v0.1 cuda 9.44s
46 Coqui XTTS-v2 cpu 9.48s
47 Qwen3-TTS 1.7B Base cuda 9.49s
48 Dia 1.6B-0626 cuda 12.27s
49 Chatterbox cpu 12.46s
50 VibeVoice Realtime 0.5B cpu 12.96s
51 OmniVoice cpu 13.20s
52 VoxCPM2 2B cpu 13.91s
53 Maya1 cuda 14.58s
54 ZipVoice 123M cpu 16.26s
55 Sesame CSM-1B cuda 17.41s
56 OuteTTS 1.0 1B cuda 23.43s
57 Miso TTS 8B cuda 24.87s
58 Mars5-TTS cpu 26.53s
59 Mars5-TTS cuda 28.21s
60 Magpie-TTS cpu 29.11s
61 Qwen3-TTS 1.7B Base cpu 30.68s
62 Fish Speech 1.5 cpu 42.69s
63 VibeVoice 1.5B cpu 45.79s
64 Zonos v0.1 cpu 49.69s
65 Parler-TTS Mini v1 cpu 52.09s
66 F5-TTS v1 cpu 58.87s
67 Sesame CSM-1B cpu 64.06s
68 ZipVoice 123M cuda 142.94s

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 Kokoro cuda 50ms
2 Piper cpu 68ms
3 MeloTTS cuda 139ms
4 NeuTTS Nano cuda 230ms
5 Pocket-TTS cpu 253ms
6 OpenVoice v2 cuda 276ms
7 NeuTTS Nano cpu 277ms
8 Kokoro cpu 334ms
9 MeloTTS cpu 464ms
10 Supertonic 3 cpu 565ms
11 OmniVoice cuda 719ms
12 Coqui XTTS-v2 cuda 859ms
13 OpenVoice v2 cpu 1.13s
14 Qwen3-TTS 1.7B (CUDA-graph) cuda 1.18s
15 MOSS-TTS-Nano cuda 2.67s
16 VoxCPM2 2B cuda 2.77s
17 Magpie-TTS cuda 3.34s
18 MOSS-TTS v1.0 cuda 3.74s
19 MOSS-TTS v1.5 cuda 3.87s
20 MOSS-TTS-Nano cpu 3.89s
21 Coqui XTTS-v2 cpu 4.47s
22 Fish Speech 1.5 cuda 4.77s
23 Zonos v0.1 cuda 5.59s
24 Qwen3-TTS 1.7B Base cuda 6.57s
25 VoxCPM2 2B cpu 8.08s
26 OmniVoice cpu 11.60s
27 ZipVoice 123M cpu 13.44s
28 OuteTTS 1.0 1B cuda 16.34s
29 Qwen3-TTS 1.7B Base cpu 22.03s
30 Magpie-TTS cpu 23.89s
31 Zonos v0.1 cpu 30.99s
32 Fish Speech 1.5 cpu 41.60s
33 ZipVoice 123M cuda 64.78s