The ULTIMATE Local AI Update Just Dropped! ACE-Step 1.5, Paints-Undo & Whisper Premium Will WOW You
ACE-Step 1.5 XL Premium generates full songs locally with the Turbo, SFT and Base models; the V8 INT8 update makes SFT and Base songs up to 1.9x faster.
Published Jun 25, 2026 · 4,089 views · Watch on YouTube
Get the app and files

ACE-Step 1.5 XL Premium: Music & Song Generator With 1-Click Install
ACE-Step 1.5 XL Premium generates full songs locally with the Turbo, SFT and Base models; the V8 INT8 update makes SFT and Base songs up to 1.9x faster.

ACE-Step 1.5 Remix, Paints-Undo & Whisper Premium: Local AI Update
ACE-Step XL 1.5 Premium remix tutorial: change lyrics and language, tune strength and melody retention, regenerate parts; a full remix takes about 33 seconds.
Chapters
- 0:00 ACE-Step XL 1.5 Premium update, remix focus, and Paints-Undo preview
- 0:49 Billie Jean remix example: checking style preservation and vocal change
- 1:16 Lower remix strength idea and Gangnam Style Korean-to-English demo
- 1:36 Hearing the Korean-to-English result and why extreme remixes sound strange
- 1:53 Stronger remix settings with higher strength and melody retention values
- 2:12 How to update ACE-Step XL 1.5: download ZIP, extract, and overwrite
- 2:37 Run Windows install/update.bat and rebuild the virtual environment if needed
- 2:56 Library tab overview: daily categories, saved generations, and metadata
- 3:18 Loading old songs from Library with lyrics, parameters, and JSON restored
- 3:33 Starting a proper remix: select the SFT model and open Advanced -) Remix
- 3:45 Uploading the source song and learning the two key remix parameters
- 4:04 Remix presets explained: different lyrics, same lyrics, medium and big change
- 4:19 Why lower melody retention lets the model generate completely new lyrics
- 4:30 Torch compile speed tip and preparing the target style caption and lyrics
- 4:42 Using Gemini in Google AI Studio with the ACE-Step lyric instruction file
- 4:59 Editing or writing lyrics and matching the vocal language accurately
- 5:15 Launching a live remix generation and measuring local generation speed
- 5:27 Live timing result: around 33 seconds for a complete remix generation
- 5:49 Use generated result as source to repair or improve selected song parts
- 6:03 Selecting remix start/stop points and regenerating only the chosen section
- 6:17 How section patching works: full remix generated, only selected part replaced
- 6:36 When to keep lyrics/style the same and when to change them for a section
- 6:47 Comparing the full output against the newly generated section preview
- 6:58 Iterative remix workflow for perfecting each part of the composition
- 7:14 LoRA training progress, future voice accuracy, and language-swap limits
- 7:27 Fast local iterations, multiple attempts, no watermark, and usable outputs
- 7:55 Final remix reminders: SFT model, proper Windows setup, and compile mode
- 8:11 Tuning dramatic changes with percentage values and fixed seed comparisons
- 8:41 Paints-Undo upgraded intro: new pipeline, faster speed, and better results
- 8:52 Download, install, and start Paints-Undo with windows_startup.bat
- 9:08 First launch model downloads plus new xFormers Triton attention support
- 9:21 GPU compatibility, Torch 2.12.1, CUDA 13, and upgrades over the original
- 9:34 Upload an image, generate the prompt, and tag it with the WD14 tagger
- 9:52 Operating steps, keyframes, Tiled VAE options, and low VRAM preparation
- 10:04 24GB vs 7GB VRAM usage and how the new memory-saving options help
- 10:27 How keyframes become the drawing video before final video generation
- 10:52 CUDA 13, Torch 2.12, Triton attention, diagnostics, and speed improvements
- 11:10 Automatic attention fallback plus Linux/cloud installer compatibility notes
- 11:27 Full generation time, possible torch compile addition, and result preview
- 11:46 Why keyframes matter, why results vary, and why not every image works
- 11:57 Fixing out-of-VRAM errors by lowering resolution and using supported ratios
- 12:14 Whisper Premium update: transcribing videos with all Whisper model options
- 12:31 Faster Whisper quality mode and the repeated sentence problem
- 12:42 Using large-v1 and repetition penalty to prevent repeated subtitle lines
- 12:54 How to tune repetition penalty carefully so transcription is not skipped
- 13:06 Example result: highly accurate subtitles generated from the new video
- 13:19 29-minute video transcribed in 1.5 minutes, 20x real-time, and closing
About this video
In this tutorial I show the newer ACE-Step XL 1.5 Premium features, especially the corrected remix workflow that was missing from the previous video. You will see how to remix songs properly, convert lyrics and language, tune remix strength and melody retention, regenerate only selected parts, use the Library metadata system, install the upgraded Paints-Undo pipeline, reduce VRAM usage, and fix repeating lines in Whisper Premium transcriptions.
Important links:
Download ACESTEP XL Premium files: [ https://www.patreon.com/posts/ace-step-1-5-xl-premium-157675060 ]
Discord: https://discord.com/invite/software-engineering-courses-secourses-772774097734074388
Google AI Studio: https://aistudio.google.com/
Previous ACE-Step tutorial: [ https://youtu.be/9C_6qNKjgpA ]
Windows requirements/setup tutorial: [ https://youtu.be/DrhUHnYfwC0 ]
Whisper Premium tutorial: [ https://youtu.be/4lAk6sf1qF8 ]
Download Whisper Premium App Files: [ https://www.patreon.com/posts/whisper-webui-premium-145395299 ]
Video chapters:
0:00 ACE-Step XL 1.5 Premium update, remix focus, and Paints-Undo preview
0:49 Billie Jean remix example: checking style preservation and vocal change
1:16 Lower remix strength idea and Gangnam Style Korean-to-English demo
1:36 Hearing the Korean-to-English result and why extreme remixes sound strange
1:53 Stronger remix settings with higher strength and melody retention values
2:12 How to update ACE-Step XL 1.5: download ZIP, extract, and overwrite
2:37 Run Windows install/update.bat and rebuild the virtual environment if needed
2:56 Library tab overview: daily categories, saved generations, and metadata
3:18 Loading old songs from Library with lyrics, parameters, and JSON restored
3:33 Starting a proper remix: select the SFT model and open Advanced -) Remix
3:45 Uploading the source song and learning the two key remix parameters
4:04 Remix presets explained: different lyrics, same lyrics, medium and big change
4:19 Why lower melody retention lets the model generate completely new lyrics
4:30 Torch compile speed tip and preparing the target style caption and lyrics
4:42 Using Gemini in Google AI Studio with the ACE-Step lyric instruction file
4:59 Editing or writing lyrics and matching the vocal language accurately
5:15 Launching a live remix generation and measuring local generation speed
5:27 Live timing result: around 33 seconds for a complete remix generation
5:49 Use generated result as source to repair or improve selected song parts
6:03 Selecting remix start/stop points and regenerating only the chosen section
6:17 How section patching works: full remix generated, only selected part replaced
6:36 When to keep lyrics/style the same and when to change them for a section
6:47 Comparing the full output against the newly generated section preview
6:58 Iterative remix workflow for perfecting each part of the composition
7:14 LoRA training progress, future voice accuracy, and language-swap limits
7:27 Fast local iterations, multiple attempts, no watermark, and usable outputs
7:55 Final remix reminders: SFT model, proper Windows setup, and compile mode
8:11 Tuning dramatic changes with percentage values and fixed seed comparisons
8:41 Paints-Undo upgraded intro: new pipeline, faster speed, and better results
8:52 Download, install, and start Paints-Undo with windows_startup.bat