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 Vaniq-Edge cuda 31ms
2 Kokoro cuda 47ms
3 Piper cpu 47ms
4 Pocket-TTS cpu 66ms
5 MeloTTS cuda 75ms
6 Vaniq-Edge cpu 82ms
7 LFM2.5-Audio 1.5B cpu 83ms
8 LFM2.5-Audio 1.5B cuda 89ms
9 Inflect-Micro v2 cuda 127ms
10 Inflect-Nano v2 cuda 128ms
11 Sopro V2 Turbo (streaming) cuda 131ms
12 StyleTTS 2 cpu 168ms
13 StyleTTS 2 cuda 175ms
14 Inflect-Micro v2 cpu 200ms
15 Audio8 TTS 0.6B (compiled) cuda 207ms
16 Kokoro cpu 226ms
17 Sopro V2 Turbo cuda 230ms
18 Inflect-Nano v2 cpu 237ms
19 NeuTTS Nano cuda 270ms
20 NeuTTS Nano cpu 303ms
21 OpenVoice v2 cuda 304ms
22 MeloTTS cpu 345ms
23 NeuTTS Air cuda 411ms
24 Supertonic 3 cpu 438ms
25 NeuTTS Air cpu 463ms
26 KittenTTS Nano 0.1 cpu 504ms
27 Scylla's Band cpu 535ms
28 Qwen3-TTS 1.7B (CUDA-graph) cuda 557ms
29 Soprano 1.1 80M cuda 557ms
30 OpenVoice v2 cpu 595ms
31 Coqui XTTS-v2 cuda 614ms
32 Soprano 1.1 80M cpu 661ms
33 Chatterbox Turbo cuda 725ms
34 F5-TTS v1 cuda 778ms
35 OmniVoice cuda 887ms
36 ZipVoice 123M cuda 1.20s
37 Chatterbox cuda 1.22s
38 LongCat-AudioDiT 1B cuda 1.24s
39 Sopro V2 Turbo (streaming) cpu 1.30s
40 VibeVoice Realtime 0.5B cuda 1.35s
41 Magpie-TTS cuda 1.61s
42 Sopro V2 Turbo cpu 1.76s
43 LongCat-AudioDiT 3.5B cuda 1.76s
44 VibeVoice 1.5B cuda 1.85s
45 IndexTTS-2 cpu 2.04s
46 Echo-TTS cuda 2.11s
47 VoxCPM2 2B cuda 2.48s
48 MOSS-TTS-Nano cuda 2.57s
49 Parler-TTS Mini v1 cuda 2.58s
50 MOSS-TTS v1.0 cuda 2.72s
51 IndexTTS-2 cuda 2.74s
52 Coqui XTTS-v2 cpu 2.75s
53 Qwen3-TTS 1.7B Base cuda 2.78s
54 MOSS-TTS-Nano cpu 2.99s
55 DramaBox cuda 2.99s
56 Fish Speech 1.5 cuda 3.06s
57 Zonos v0.1 cuda 3.50s
58 Chatterbox Turbo cpu 3.62s
59 Sesame CSM-1B cuda 3.73s
60 Audio8 TTS Preview 0.6B cuda 3.96s
61 VibeVoice Realtime 0.5B cpu 4.74s
62 Maya1 cuda 4.75s
63 Dia 1.6B-0626 cuda 4.75s
64 Audio8 TTS Preview 0.1B cuda 4.87s
65 Qwen3-TTS 0.6B CustomVoice cuda 4.94s
66 Chatterbox cpu 5.32s
67 Miso TTS 8B cuda 5.48s
68 MOSS-TTS v1.5 cuda 5.62s
69 Magpie-TTS cpu 7.12s
70 OmniVoice cpu 7.50s
71 VibeVoice 7B cuda 7.64s
72 VoxCPM2 2B cpu 8.44s
73 OuteTTS 1.0 1B cuda 10.20s
74 ZipVoice 123M cpu 10.63s
75 Qwen3-TTS 1.7B Base cpu 11.69s
76 Qwen3-TTS 0.6B CustomVoice cpu 13.53s
77 Audio8 TTS Preview 0.6B cpu 13.96s
78 Audio8 TTS Preview 0.1B cpu 14.47s
79 Fish Speech 1.5 cpu 15.07s
80 Parler-TTS Mini v1 cpu 16.48s
81 Mars5-TTS cpu 19.20s
82 Mars5-TTS cuda 19.75s
83 Zonos v0.1 cpu 19.96s
84 VibeVoice 1.5B cpu 20.22s
85 Sesame CSM-1B cpu 20.89s
86 F5-TTS v1 cpu 45.35s
87 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 Vaniq-Edge cuda 27ms
2 Kokoro cuda 59ms
3 Piper cpu 89ms
4 LFM2.5-Audio 1.5B cpu 89ms
5 Pocket-TTS cpu 94ms
6 LFM2.5-Audio 1.5B cuda 103ms
7 MeloTTS cuda 107ms
8 Inflect-Nano v2 cuda 117ms
9 Inflect-Micro v2 cuda 124ms
10 Sopro V2 Turbo (streaming) cuda 138ms
11 Vaniq-Edge cpu 170ms
12 StyleTTS 2 cuda 188ms
13 StyleTTS 2 cpu 198ms
14 Inflect-Nano v2 cpu 247ms
15 NeuTTS Nano cuda 268ms
16 NeuTTS Nano cpu 317ms
17 OpenVoice v2 cuda 353ms
18 Inflect-Micro v2 cpu 361ms
19 Sopro V2 Turbo cuda 370ms
20 Kokoro cpu 380ms
21 NeuTTS Air cuda 420ms
22 Audio8 TTS 0.6B (compiled) cuda 434ms
23 NeuTTS Air cpu 506ms
24 Scylla's Band cpu 556ms
25 Supertonic 3 cpu 586ms
26 MeloTTS cpu 656ms
27 F5-TTS v1 cuda 735ms
28 OmniVoice cuda 773ms
29 Sopro V2 Turbo (streaming) cpu 814ms
30 KittenTTS Nano 0.1 cpu 843ms
31 LongCat-AudioDiT 1B cuda 971ms
32 Coqui XTTS-v2 cuda 1.15s
33 OpenVoice v2 cpu 1.18s
34 Qwen3-TTS 1.7B (CUDA-graph) cuda 1.30s
35 Chatterbox Turbo cuda 1.30s
36 Soprano 1.1 80M cuda 1.33s
37 Soprano 1.1 80M cpu 1.43s
38 LongCat-AudioDiT 3.5B cuda 1.49s
39 Echo-TTS cuda 2.20s
40 Chatterbox cuda 2.24s
41 VibeVoice Realtime 0.5B cuda 2.64s
42 Sopro V2 Turbo cpu 2.80s
43 DramaBox cuda 3.42s
44 Magpie-TTS cuda 3.45s
45 VibeVoice 1.5B cuda 3.47s
46 VoxCPM2 2B cuda 3.59s
47 VibeVoice 7B cuda 3.78s
48 MOSS-TTS v1.0 cuda 3.94s
49 IndexTTS-2 cuda 4.16s
50 IndexTTS-2 cpu 4.23s
51 MOSS-TTS-Nano cpu 4.47s
52 MOSS-TTS-Nano cuda 5.23s
53 MOSS-TTS v1.5 cuda 5.89s
54 Parler-TTS Mini v1 cuda 6.04s
55 Qwen3-TTS 1.7B Base cuda 6.47s
56 Coqui XTTS-v2 cpu 6.65s
57 Fish Speech 1.5 cuda 7.07s
58 Zonos v0.1 cuda 7.78s
59 Chatterbox Turbo cpu 8.06s
60 Audio8 TTS Preview 0.6B cuda 8.70s
61 VibeVoice Realtime 0.5B cpu 9.36s
62 Qwen3-TTS 0.6B CustomVoice cuda 10.10s
63 Chatterbox cpu 10.32s
64 Sesame CSM-1B cuda 10.43s
65 Maya1 cuda 10.54s
66 VoxCPM2 2B cpu 11.56s
67 OmniVoice cpu 13.09s
68 Audio8 TTS Preview 0.1B cuda 14.39s
69 OuteTTS 1.0 1B cuda 20.98s
70 Magpie-TTS cpu 21.02s
71 Miso TTS 8B cuda 22.52s
72 Qwen3-TTS 1.7B Base cpu 22.87s
73 Audio8 TTS Preview 0.1B cpu 24.24s
74 Qwen3-TTS 0.6B CustomVoice cpu 24.54s
75 Mars5-TTS cuda 25.09s
76 Mars5-TTS cpu 26.10s
77 VibeVoice 1.5B cpu 29.51s
78 Fish Speech 1.5 cpu 31.14s
79 Audio8 TTS Preview 0.6B cpu 31.27s
80 Parler-TTS Mini v1 cpu 42.46s
81 Zonos v0.1 cpu 45.49s
82 Dia 1.6B-0626 cuda 51.96s
83 F5-TTS v1 cpu 58.82s
84 Sesame CSM-1B cpu 66.45s
85 Maya1 cpu 97.77s
86 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 Vaniq-Edge cuda 33ms
2 LFM2.5-Audio 1.5B cuda 79ms
3 LFM2.5-Audio 1.5B cpu 84ms
4 Pocket-TTS cpu 109ms
5 Kokoro cuda 116ms
6 Sopro V2 Turbo (streaming) cuda 129ms
7 Inflect-Micro v2 cuda 136ms
8 MeloTTS cuda 153ms
9 Inflect-Nano v2 cuda 167ms
10 Piper cpu 239ms
11 NeuTTS Nano cuda 265ms
12 NeuTTS Nano cpu 322ms
13 StyleTTS 2 cpu 328ms
14 StyleTTS 2 cuda 374ms
15 NeuTTS Air cuda 423ms
16 NeuTTS Air cpu 458ms
17 OpenVoice v2 cuda 485ms
18 Vaniq-Edge cpu 568ms
19 Inflect-Nano v2 cpu 593ms
20 OmniVoice cuda 714ms
21 Sopro V2 Turbo (streaming) cpu 822ms
22 Inflect-Micro v2 cpu 877ms
23 F5-TTS v1 cuda 1.03s
24 Audio8 TTS 0.6B (compiled) cuda 1.18s
25 Scylla's Band cpu 1.27s
26 Kokoro cpu 1.27s
27 Sopro V2 Turbo cuda 1.27s
28 LongCat-AudioDiT 1B cuda 1.29s
29 Supertonic 3 cpu 1.45s
30 MeloTTS cpu 1.76s
31 Echo-TTS cuda 2.20s
32 LongCat-AudioDiT 3.5B cuda 2.20s
33 KittenTTS Nano 0.1 cpu 2.46s
34 OpenVoice v2 cpu 3.11s
35 Chatterbox Turbo cuda 3.12s
36 Qwen3-TTS 1.7B (CUDA-graph) cuda 3.30s
37 Soprano 1.1 80M cuda 3.70s
38 Soprano 1.1 80M cpu 4.27s
39 Chatterbox cuda 4.68s
40 Coqui XTTS-v2 cuda 4.82s
41 DramaBox cuda 5.19s
42 VibeVoice Realtime 0.5B cuda 7.86s
43 MOSS-TTS v1.0 cuda 8.40s
44 VibeVoice 7B cuda 8.77s
45 Magpie-TTS cuda 9.64s
46 MOSS-TTS-Nano cuda 10.45s
47 VibeVoice 1.5B cuda 10.63s
48 Sopro V2 Turbo cpu 10.83s
49 VoxCPM2 2B cuda 11.06s
50 IndexTTS-2 cpu 11.13s
51 MOSS-TTS-Nano cpu 11.83s
52 MOSS-TTS v1.5 cuda 11.88s
53 IndexTTS-2 cuda 12.58s
54 Fish Speech 1.5 cuda 17.39s
55 Parler-TTS Mini v1 cuda 18.07s
56 Sesame CSM-1B cuda 18.23s
57 Chatterbox Turbo cpu 19.26s
58 Qwen3-TTS 1.7B Base cuda 19.42s
59 Dia 1.6B-0626 cuda 22.29s
60 Audio8 TTS Preview 0.6B cuda 24.09s
61 Coqui XTTS-v2 cpu 25.28s
62 Zonos v0.1 cuda 25.58s
63 VibeVoice Realtime 0.5B cpu 26.34s
64 Chatterbox cpu 28.27s
65 Maya1 cuda 29.31s
66 Qwen3-TTS 0.6B CustomVoice cuda 30.99s
67 VoxCPM2 2B cpu 32.23s
68 OmniVoice cpu 34.51s
69 Audio8 TTS Preview 0.1B cuda 42.41s
70 OuteTTS 1.0 1B cuda 50.39s
71 Mars5-TTS cpu 50.94s
72 Mars5-TTS cuda 51.19s
73 Qwen3-TTS 1.7B Base cpu 63.06s
74 Audio8 TTS Preview 0.1B cpu 74.04s
75 Miso TTS 8B cuda 75.59s
76 F5-TTS v1 cpu 77.80s
77 Sesame CSM-1B cpu 78.74s
78 Qwen3-TTS 0.6B CustomVoice cpu 79.10s
79 VibeVoice 1.5B cpu 85.74s
80 Magpie-TTS cpu 87.47s
81 Fish Speech 1.5 cpu 96.22s
82 Audio8 TTS Preview 0.6B cpu 101.71s
83 Parler-TTS Mini v1 cpu 140.05s
84 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 Vaniq-Edge cuda 27ms
2 Kokoro cuda 65ms
3 LFM2.5-Audio 1.5B cpu 78ms
4 LFM2.5-Audio 1.5B cuda 91ms
5 Pocket-TTS cpu 92ms
6 Piper cpu 93ms
7 MeloTTS cuda 108ms
8 Inflect-Nano v2 cuda 113ms
9 Inflect-Micro v2 cuda 114ms
10 Sopro V2 Turbo (streaming) cuda 130ms
11 Vaniq-Edge cpu 180ms
12 Inflect-Nano v2 cpu 255ms
13 NeuTTS Nano cuda 258ms
14 NeuTTS Nano cpu 297ms
15 StyleTTS 2 cpu 309ms
16 StyleTTS 2 cuda 323ms
17 Inflect-Micro v2 cpu 370ms
18 OpenVoice v2 cuda 387ms
19 NeuTTS Air cuda 417ms
20 Kokoro cpu 452ms
21 NeuTTS Air cpu 456ms
22 Sopro V2 Turbo cuda 571ms
23 Supertonic 3 cpu 663ms
24 Audio8 TTS 0.6B (compiled) cuda 673ms
25 OmniVoice cuda 694ms
26 MeloTTS cpu 743ms
27 Sopro V2 Turbo (streaming) cpu 778ms
28 Scylla's Band cpu 814ms
29 F5-TTS v1 cuda 836ms
30 KittenTTS Nano 0.1 cpu 1.01s
31 LongCat-AudioDiT 1B cuda 1.26s
32 Chatterbox Turbo cuda 1.33s
33 OpenVoice v2 cpu 1.39s
34 Soprano 1.1 80M cuda 1.49s
35 Soprano 1.1 80M cpu 1.62s
36 Qwen3-TTS 1.7B (CUDA-graph) cuda 1.66s
37 LongCat-AudioDiT 3.5B cuda 1.74s
38 Coqui XTTS-v2 cuda 1.93s
39 Echo-TTS cuda 2.09s
40 Chatterbox cuda 2.31s
41 VibeVoice Realtime 0.5B cuda 3.23s
42 DramaBox cuda 3.52s
43 MOSS-TTS-Nano cuda 3.67s
44 IndexTTS-2 cpu 4.29s
45 Magpie-TTS cuda 4.37s
46 VibeVoice 7B cuda 4.48s
47 IndexTTS-2 cuda 4.64s
48 MOSS-TTS v1.0 cuda 4.90s
49 Sopro V2 Turbo cpu 4.92s
50 MOSS-TTS-Nano cpu 5.00s
51 VibeVoice 1.5B cuda 5.30s
52 Parler-TTS Mini v1 cuda 5.87s
53 VoxCPM2 2B cuda 6.05s
54 Fish Speech 1.5 cuda 6.90s
55 MOSS-TTS v1.5 cuda 7.04s
56 Chatterbox Turbo cpu 8.76s
57 Zonos v0.1 cuda 9.44s
58 Coqui XTTS-v2 cpu 9.48s
59 Qwen3-TTS 1.7B Base cuda 9.49s
60 Audio8 TTS Preview 0.6B cuda 10.83s
61 Dia 1.6B-0626 cuda 12.27s
62 Chatterbox cpu 12.46s
63 VibeVoice Realtime 0.5B cpu 12.96s
64 OmniVoice cpu 13.20s
65 VoxCPM2 2B cpu 13.91s
66 Maya1 cuda 14.58s
67 Qwen3-TTS 0.6B CustomVoice cuda 15.33s
68 ZipVoice 123M cpu 16.26s
69 Sesame CSM-1B cuda 17.41s
70 Audio8 TTS Preview 0.1B cuda 18.41s
71 OuteTTS 1.0 1B cuda 23.43s
72 Miso TTS 8B cuda 24.87s
73 Mars5-TTS cpu 26.53s
74 Mars5-TTS cuda 28.21s
75 Magpie-TTS cpu 29.11s
76 Qwen3-TTS 1.7B Base cpu 30.68s
77 Audio8 TTS Preview 0.1B cpu 34.22s
78 Qwen3-TTS 0.6B CustomVoice cpu 35.06s
79 Fish Speech 1.5 cpu 42.69s
80 VibeVoice 1.5B cpu 45.79s
81 Audio8 TTS Preview 0.6B cpu 47.85s
82 Zonos v0.1 cpu 49.69s
83 Parler-TTS Mini v1 cpu 52.09s
84 F5-TTS v1 cpu 58.87s
85 Sesame CSM-1B cpu 64.06s
86 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 Sopro V2 Turbo (streaming) cuda 132ms
4 MeloTTS cuda 139ms
5 NeuTTS Nano cuda 230ms
6 Pocket-TTS cpu 253ms
7 OpenVoice v2 cuda 276ms
8 NeuTTS Nano cpu 277ms
9 Kokoro cpu 334ms
10 Sopro V2 Turbo cuda 346ms
11 Audio8 TTS 0.6B (compiled) cuda 398ms
12 MeloTTS cpu 464ms
13 Supertonic 3 cpu 565ms
14 OmniVoice cuda 719ms
15 Sopro V2 Turbo (streaming) cpu 844ms
16 Coqui XTTS-v2 cuda 859ms
17 OpenVoice v2 cpu 1.13s
18 Qwen3-TTS 1.7B (CUDA-graph) cuda 1.18s
19 Sopro V2 Turbo cpu 2.48s
20 MOSS-TTS-Nano cuda 2.67s
21 VoxCPM2 2B cuda 2.77s
22 Magpie-TTS cuda 3.34s
23 MOSS-TTS v1.0 cuda 3.74s
24 MOSS-TTS v1.5 cuda 3.87s
25 MOSS-TTS-Nano cpu 3.89s
26 Coqui XTTS-v2 cpu 4.47s
27 Fish Speech 1.5 cuda 4.77s
28 Zonos v0.1 cuda 5.59s
29 Audio8 TTS Preview 0.6B cuda 6.49s
30 Qwen3-TTS 1.7B Base cuda 6.57s
31 Qwen3-TTS 0.6B CustomVoice cuda 8.04s
32 VoxCPM2 2B cpu 8.08s
33 OmniVoice cpu 11.60s
34 Audio8 TTS Preview 0.1B cuda 11.91s
35 ZipVoice 123M cpu 13.44s
36 OuteTTS 1.0 1B cuda 16.34s
37 Audio8 TTS Preview 0.1B cpu 16.73s
38 Qwen3-TTS 0.6B CustomVoice cpu 19.71s
39 Qwen3-TTS 1.7B Base cpu 22.03s
40 Magpie-TTS cpu 23.89s
41 Audio8 TTS Preview 0.6B cpu 26.79s
42 Zonos v0.1 cpu 30.99s
43 Fish Speech 1.5 cpu 41.60s
44 ZipVoice 123M cuda 64.78s