Patreon
Home / Videos / I Trained My Own AI Voice on My PC, and Users Rate It Above

I Trained My Own AI Voice on My PC, and Users Rate It Above ElevenLabs

Ultimate TTS full Windows tutorial covering voice cloning, automatic take selection, multilingual speech and OmniVoice full fine-tuning in 600+ languages.

Published Oct 06, 2026 · 9,320 views · Watch on YouTube

Get the app and files

Ultimate TTS Tutorial: Voice Training & Cloning in 600+ Languages

Ultimate TTS Tutorial: Voice Training & Cloning in 600+ Languages

Ultimate TTS full Windows tutorial covering voice cloning, automatic take selection, multilingual speech and OmniVoice full fine-tuning in 600+ languages.

Ultimate TTS: Voice Cloning & Training App With IndexTTS 2.5 & OmniVoice

Ultimate TTS: Voice Cloning & Training App With IndexTTS 2.5 & OmniVoice

Ultimate Text to Speech Generator clones any voice from a short clip and trains your own voice; IndexTTS 2.5 made 31 s of speech in 18 s on an RTX 5090.

Chapters

  1. 0:00 My trained AI voice
  2. 0:38 600+ supported languages
  3. 1:19 Windows and cloud options
  4. 1:46 Your first voice clone
  5. 2:10 Generate and listen
  6. 3:18 Download ZIP and guides
  7. 3:46 Windows Requirements
  8. 4:28 Choose an install path
  9. 4:40 Extract the package
  10. 4:54 Run the installer
  11. 5:23 PyTorch, CUDA and wheels
  12. 5:54 Start the application
  13. 6:19 Models, voices, GPU presets
  14. 7:04 First load vs warm runs
  15. 7:46 Players and saved outputs
  16. 8:34 IndexTTS English example
  17. 9:02 AuK English example
  18. 9:52 Reference library and files
  19. 10:48 Select a reference range
  20. 11:38 Microphone recording
  21. 11:53 Reference text and transcription
  22. 13:19 Clone and Auto voice
  23. 14:38 Voice design tags
  24. 15:26 Language selection
  25. 15:40 Preview and Smart sentences
  26. 16:32 Section-length study
  27. 17:01 Sentence and section pauses
  28. 18:08 Pause tags
  29. 19:04 Check and set pronunciation
  30. 19:38 Save and reuse dictionary
  31. 21:10 Expressive tags
  32. 21:44 Keep which take
  33. 21:58 Fewest word errors
  34. 22:12 Rank takes and set checks
  35. 22:42 Ranking and fallback result
  36. 23:49 Whole-text candidates
  37. 24:36 60-line selection study
  38. 25:05 139-line Auto voice study
  39. 25:35 First vs selected take
  40. 26:42 Matching-language references
  41. 27:26 Arabic voice example
  42. 28:24 Chinese voice example
  43. 29:04 French voice example
  44. 29:46 German voice example
  45. 30:40 Italian voice example
  46. 31:22 Japanese voice example
  47. 32:03 Korean voice example
  48. 32:58 Polish voice example
  49. 33:38 Spanish voice example
  50. 33:49 Cross-language cloning
  51. 35:48 Speaking-rate comparison
  52. 36:49 Target duration modes
  53. 37:39 Best quality: 32 steps
  54. 38:25 Fast: 16 steps
  55. 38:54 Seeds and advanced sampling
  56. 39:44 WAV, MP3 and filenames
  57. 40:01 Reference and subtitle export
  58. 40:16 Inspect the exported files
  59. 41:10 Recent outputs as references
  60. 41:52 Audio tuning comparison
  61. 43:09 Tuning and edge silence
  62. 43:30 SRT import and cue timing
  63. 44:22 Create a narrated MP4
  64. 45:08 Batch inputs and filenames
  65. 45:43 Shared and per-file voices
  66. 46:07 Batch execution and progress
  67. 46:47 Inspect batch results
  68. 48:00 Save a personal preset
  69. 48:18 Load settings and reference
  70. 49:26 Recent settings and theme
  71. 49:45 Select GPU and memory tier
  72. 50:02 BF16 and ConvRot INT8
  73. 50:16 Measured memory use
  74. 50:30 Measured synthesis time
  75. 50:44 Compare both variants
  76. 52:25 Section batch size
  77. 52:39 Load models and find files
  78. 53:03 VRAM benchmark and tier cap
  79. 53:56 Train a new OmniVoice
  80. 54:15 Source recordings and text
  81. 54:44 Reserve evaluation sessions
  82. 55:00 Dataset folder and scan
  83. 55:37 Transcript and alignment
  84. 55:57 Segment duration
  85. 56:13 Padding and caption cleanup
  86. 56:28 Quality filters
  87. 56:47 Prepare and inspect clips
  88. 58:02 Speaker and transcript audit
  89. 58:50 Review flagged clips
  90. 59:21 Correct words and boundaries
  91. 60:20 Build the OmniVoice cache
  92. 60:47 Dataset and training name
  93. 61:06 Full fine-tuning
  94. 61:33 Full, LoRA and DoRA
  95. 61:50 Held-out voice comparison
  96. 62:05 Epochs and Training plan
  97. 62:22 Token budget and accumulation
  98. 62:39 Learning rate and memory
  99. 63:14 Checkpoints and train state
  100. 63:30 Continue vs weights-only
  101. 63:47 Evaluation and final test
  102. 64:18 Fluency analysis and filters
  103. 65:16 Start training
  104. 65:49 Speed, loss and validation
  105. 66:18 Hear a training sample
  106. 66:45 Stop and resume controls
  107. 67:05 800-update completion
  108. 67:39 Checkpoint analysis
  109. 67:55 Automatic speech comparison
  110. 68:15 Reference audition
  111. 68:37 Speaking-rate calibration
  112. 68:53 Save an INT8 voice copy
  113. 69:46 Load the trained preset
  114. 70:14 Generate held-out speech
  115. 71:12 Base vs trained voice
  116. 72:15 Configure Checkpoint Grid
  117. 73:24 Hear the six grid cells
  118. 74:34 Trained Auto voice
  119. 75:22 Trained INT8 voice
  120. 76:13 Updates, backup and help
  121. 77:01 Your next voice

About this video

Train your own AI voice with Ultimate TTS. This complete Windows tutorial covers voice cloning, automatic take selection, multilingual speech and OmniVoice full fine-tuning. The app also supports RunPod, SimplePod, Massed Compute and local Linux.

Links:

Download Ultimate TTS and read the source post: [ https://www.patreon.com/SECourses/posts/ultimate-text-to-139297407 ]

Windows Requirements tutorial: [ https://youtu.be/DrhUHnYfwC0 ]

You will learn how to install the app, clone an included voice, prepare references, compare takes, control pronunciation and timing, export subtitles and MP4s, run batches, audit a dataset and deploy a newly trained voice.

Main topics include OmniVoice, IndexTTS 2.5, AuK, 9 language examples, BF16 and ConvRot INT8, speaker similarity, Whisper checks, Full/LoRA/DoRA training, checkpoint grids and voice presets.

Use the chapters to jump directly to each control, example and training step.

Chapters:

0:00 My trained AI voice

0:38 600+ supported languages

1:19 Windows and cloud options

1:46 Your first voice clone

2:10 Generate and listen

3:18 Download ZIP and guides

3:46 Windows Requirements

4:28 Choose an install path

4:40 Extract the package

4:54 Run the installer

5:23 PyTorch, CUDA and wheels

5:54 Start the application

6:19 Models, voices, GPU presets

7:04 First load vs warm runs

7:46 Players and saved outputs

8:34 IndexTTS English example

9:02 AuK English example

9:52 Reference library and files

10:48 Select a reference range

11:38 Microphone recording

11:53 Reference text and transcription

13:19 Clone and Auto voice

14:38 Voice design tags

15:26 Language selection

15:40 Preview and Smart sentences

16:32 Section-length study

17:01 Sentence and section pauses

18:08 Pause tags

19:04 Check and set pronunciation

19:38 Save and reuse dictionary

21:10 Expressive tags

21:44 Keep which take