#
Applio Lightning Ai
Last update: August 8, 2025

#
Introduction
This is a cloud-based alternative to run Applio, RVC Fork, only for people who don't have a good PC GPU, via the Lightning.AI Service. With a Web User Interface.
It's a great alternative for training RVC voice models through the cloud, since it has the best GPUs with tons of VRAM.
Lightning.AI Service
Check the Lightning.AI Glossary for more info on Free Tier, Limits, Verification, Pricing and other things.
#
Pros & Cons
The pros & cons are subjective to your necessities.
- Has good GPU's.
- Has lots of VRAM
- TensorBoard included.
- You can leave training unsupervised.
- Takes some time to set up.
- Needs a phone number.
- Low/Decent GPU time depending on what GPU you choose.
- 2-3 Day verification wait time.
#
Create an Account
#
1. Set up account.
- First make an account with Lightning Ai

- Make sure you verify yourself with a phone number. Once you've done that you will get an email that looks like this:

You will need to wait 2-3 business days to become fully verified
- Once you are verified Lightning Ai will send you a email that conatins this:

#
Studio Setup & Installation
#
2. Access the Notebook
- After creating your Lightning.AI account, open the Applio Notebook and Clone it.

#
3. Activate/Switch GPU
- If you aren't on a GPU environment by default, you must switch to a GPU environment. This is crucial for performance.
- On the right-hand lateral menu, click on Studio Environment (the processor icon).
- Click Switch To GPU, select an available GPU, and wait for the environment to restart.
Here is a list of how long you can use each GPU before running out of Free credits.
- 75 hours monthly of T4 16gb
- 31 hours monthly of L4 24gb
- 15 hours monthly of L40 48gb
#
4. Clone Repository and Install Dependencies
- Run the first code cell. This will download the latest version of the realtime voice changer and install necessary dependencies.
- This step may take a few minutes to complete. It will print "Installed!" when finished.
#
Tunnels
#
5. Launch the Server via Tunnels
This final code cell is the most important one—it starts the voice changer's server and uses a "tunneling" service to create a secure, public web address (URL) for you to access it from your own computer.
Navigate to the third code cell, titled "Start Server using Tunnels". This cell boots up the Wokada Deiteris Fork application inside your Lightning.AI Studio.
Select a Tunnel: A tunnel securely exposes the application running in your private cloud environment to the public internet. The notebook gives you five different services to do this. Choose one from the
Tunnel
code menu in the code cell.Port Viewer (Recommended & Default method)
- How it works: This is a built-in Lightning.AI feature. It's the most straightforward method as it doesn't require any external accounts or tokens.
- Steps:
- Select "Port Viewer" from the
Tunnel
code. - Click the + at the bottom of the right tab, click on Web Apps and install Port Viewer.
- Run the code cell. Wait for the output to show that the server is listening.
- In the right-hand sidebar of the Lightning.AI interface, click the Web Apps tab.
- Click on Port Viewer and then click Add a new port.
- Enter
18888
as the Port Number and optionally give it a name (e.g., "Voice Changer"). - Click your Port in Port Viewer, you can also click Open to open it in an external tab.
- You can optionally go back to the Jupyter session in the right-hand sidebar of the Lightning.AI interface, to check if any error appears in the code output.
- Select "Port Viewer" from the
Ngrok (Fast, Popular & Reliable)
- How it works: Ngrok is a popular service that creates secure tunnels. It requires a free account and an authentication token. It has a 1GB Bandwidth Free Monthly Limit https://ngrok.com/docs/pricing-limits/free-plan-limits/.
- Steps:
- Go to the Ngrok Dashboard to get your free authtoken.
- In the notebook cell, paste your token into the
Token
field, replacing'Ngrok | Horizon TOKEN'
. - (Optional) To potentially reduce latency, select a geographical
Region
from the list of parameter options next to it, that is closest to you. - Run the cell. The public Ngrok URL (ending in
ngrok.io
) will be printed in the output once the server is ready. Click on it to access the UI.
Cloudflare (Easy, No Account Needed)
- How it works: This option uses Cloudflare's free
trycloudflare
service. It's very easy to use as it requires no accounts or tokens. - Steps:
- Select "Cloudflare" from the
Tunnel
code. - Run the cell. The script will automatically download the necessary tools. After a few moments, a public URL (ending in
trycloudflare.com
) will be printed in the output. Click it to open the interface.
- Select "Cloudflare" from the
- How it works: This option uses Cloudflare's free
LocalTunnel (No Account, Password Protected)
- How it works: LocalTunnel is another free service that doesn't require an account. For security, it generates a unique URL that is protected by a password.
- Steps:
- Select "LocalTunnel" from the
Tunnel
code. - Run the cell.
- The output will display two key pieces of information: the public URL (ending in
loca.lt
) and aLocal Tunnel Password
below it. - Click the URL. A new page will ask for a password.
- Copy the password from the notebook output and paste it into the password prompt in your browser to access the voice changer.
- Select "LocalTunnel" from the
Horizon (Fast, Requires Account & ID)
- How it works: Horizon is another tunneling service that requires a free account and a personal ID for authentication.
- Steps:
- Go to the Horizon Dashboard and sign up. On the second step of the setup, you will see a command like
hrzn login YOUR_ID
. Copy thatYOUR_ID
part. - In the notebook cell, paste this ID into the
Token
field. - Run the cell. The first time you use it, the output may ask you to authorize the connection by clicking a link (
https://hrzn.run/dashboard/settings/cli-token-requests/...
). Click this link and approve the request in your Horizon dashboard. - The public Horizon URL (ending in
hrzn.run
) will then be printed in the output. Click it to access the UI.
- Go to the Horizon Dashboard and sign up. On the second step of the setup, you will see a command like
After configuring your chosen tunnel, run the cell. The first time you run it, it might download the necessary files, which might take a minute or two.
Once the setup is complete, the output will display a message with your public URL. Click this link to open the Applio interface and start using the program.
Note:
The server runs in the foreground. If you stop the cell or close the Lightning.AI site, the server will shut down. Keep the cell running to use the program.
#
Server Setup
#
6. Accessing Files.
- To upload a dataset, upload audio or anything else find the
Teamspace Drive
button on the right and click it.
The path to Applio is Studio > this_studio > Applio > Applio
Once you're there you can just drag and drop files.
To download files click on the file then click the three dots on the right of it and click download

#
7. Opening the TensorBoard.
- Find the TensorBoard icon on the right side bar and click it.

- Once you've clicked it, Start it.

- Once you've done that it will open the TensorBoard. you can open it externally in another tab/window via clicking Open. To learn how to use it go here
#
8. Opening the notebook.
- If you want to go back to the notebook simply click on the
Jupyter
icon on the right.

#
Usage
Now that you have the web interface running via Lightning.AI, the rest of the process is identical to using a local installation.
For all subsequent steps, including application settings and model usage, please continue by following the Local PC guide.
Continue with the Local PC Guide
#
Maintenance
#
Deleting Everything
If you need to update Applio or start fresh, you can run the final cell in the notebook, "Delete everything". This will remove all downloaded files and configurations from your persistent storage, allowing for a clean installation by re-following the notebook with perhaps a changed branch variable.