What is VoltageGPU API?

VoltageGPU provides two APIs: a GPU Pods API for deploying and managing GPU instances, and an AI Inference API for running 140+ AI models including DeepSeek-R1, Llama 3, and FLUX.

Is VoltageGPU API OpenAI compatible?

Yes, the VoltageGPU AI Inference API is fully OpenAI-compatible. You can use existing OpenAI SDKs by simply changing the base URL to https://api.voltagegpu.com/v1

What AI models are available?

VoltageGPU offers 140+ AI models including DeepSeek-R1, DeepSeek-V3, Llama 3.3 70B, Qwen 2.5, Mixtral, FLUX.1, Stable Diffusion, and many more for chat, embeddings, image and video generation.

Documentation

Restores

The Restore feature in VoltageGPU allows you to restore backups of your data into running pods. Learn how to recover your training checkpoints, models, and datasets.

How to Restore Backups

Option 1: Restore Backup into Running Pod

This option allows you to select a backup from a running pod and restore it to a target pod.

Step 1: Navigate to the "Your Pod" page of the pod that has backup configuration and backups
Step 2: Click the "Backups" tab
Step 3: Click the "Restore" button
Step 4: Configure the restore modal:
- Target Pod: The pod where the selected backup will be restored
- Restore Path: The directory path in the target pod where the backup will be extracted

📝 Important

The restore path must be a subdirectory of the target pod's volume path.

Option 2: Configure Restore When Deploying a Pod

This option allows you to select a backup during the pod deployment process.

Step 1: Go to "Browse Pods" and click the "Rent Now" button
Step 2: At the bottom, locate the "Restore" section
Step 3: Click the "Select a Backup" button
Step 4: Configure the restore settings in the modal:
- Select from currently running pods
- Once a pod is selected, its available backups will be displayed
- Choose the backup you want to restore
- Restore Path: The directory in the deploying pod where the backup will be extracted

✅ Pro Tip

Restoring during deployment is perfect for quickly spinning up a new pod with all your previous work already in place.

Restore Path Configuration

The restore path is the directory where the backup ZIP file will be extracted. It must be a subdirectory of the pod's local volume mount path (default: /root).

Valid Examples ✅

/root                       # Entire volume mount (restores everything)
/root/models                # Subdirectory of /root volume mount
/root/checkpoints          # Training checkpoints directory
/root/project/data         # Nested subdirectory of /root

Invalid Examples ❌

/home/user/documents       # Not under /root volume mount path
/tmp/backup                # Temporary directory, not persistent storage
/var/log                   # System directory outside volume mount
/mnt/external              # External volumes (use /root for restores)

⚠️ Why This Matters

Valid paths are subdirectories within your local volume mount (/root), ensuring data persists and is accessible for restore operations. Invalid paths point to non-persistent storage or system directories, which can cause restore failures.

View Restore History

To view the history of all restore operations for a pod:

Navigate to the "Your Pod" page
Click the "Restores" tab
View a chronological list of all restore operations

The restore history shows:

Timestamp of each restore operation
Source backup ID
Restore path used
Status (success/failed)

Best Practices

Verify Backup Contents

Before restoring, check the backup details to ensure it contains the data you need.

Use Specific Paths

Restore to specific directories rather than /root to avoid overwriting existing files.

Check Available Space

Ensure the target pod has sufficient storage space for the restored data.

Test After Restore

Verify that restored files are accessible and your applications work correctly.

Frequently Asked Questions

Can I restore to a different pod than the original?

Yes! Backup restoration is not tied to specific pods. You can restore any backup to any pod with sufficient storage space.

What happens to existing files in the restore path?

Files from the backup will be extracted to the restore path. Existing files with the same names will be overwritten.

How long does a restore take?

Restore time depends on the backup size and network speed. Small backups restore in seconds, while large backups may take several minutes.

Can I restore while the pod is running workloads?

Yes, but be cautious. Restoring while applications are running may cause conflicts if they're accessing the same files being restored.

What if my restore fails?

Check the restore history for error details. Common issues include insufficient storage space or invalid restore paths. The original backup remains intact.

Can I restore multiple backups to the same pod?

Yes, you can restore multiple backups sequentially. Use different restore paths to avoid overwriting data from previous restores.

Ready to Restore Your Data?

Recover your training checkpoints and models in minutes.

View Your Pods Configure Backups