Privacy Policy

Last updated: January 19, 2026

1. Introduction

SimpleTuner.io ("we", "our", "us") respects your privacy. This Privacy Policy explains how we collect, use, and protect your information when you use our cloud GPU training service ("Service").

2. Information We Collect

Account Information

When you create an account, we collect:

• Email address
• Name (optional)
• Organization name (optional)

Account authentication is handled by Clerk. Please review Clerk's Privacy Policy for details on their data handling.

Payment Information

Payment processing is handled by Dodo Payments as our Merchant of Record. We do not store your credit card details. Dodo Payments collects:

• Billing name and address
• Payment method details
• Transaction history

Please review Dodo Payments' Privacy Policy for details.

Job Data

When you use the Service, we collect:

• Training configuration files (model settings, hyperparameters)
• Job metadata (status, duration, timestamps)
• Error messages and logs (for troubleshooting)
• Content moderation results (compliance status, rejection reasons)

We retain training data only as needed to provide the Service:

• Your training datasets (staged only during the upload window and job execution by default)
• Managed storage datasets (stored in your organization's Cloudflare R2 bucket until your TTL expires, default 90 days, or you delete them; may be deleted for non-payment of storage fees)
• Your trained model weights (available for download during a short post-job window)
• Your training outputs (available for download during a short post-job window)

We do not sell, transfer, or use your training data for any purpose other than providing the Service.

Anonymized Benchmark Data

During training, SimpleTuner emits performance metrics that we store without any link to your user ID, organization, or job. This data powers our runtime estimation system, helping all users get accurate cost predictions before submitting jobs.

Training Step Benchmarks - we collect:

• Model recipe (e.g., "FLUX LoRA", "SDXL full fine-tune")
• Hardware profile (e.g., "RTX 4090", "H100")
• Resolution bucket (rounded to nearest 0.5 megapixels)
• Batch size and gradient accumulation steps
• Precision (fp16, bf16) and quantization level (none, int8, nf4)
• Average step time in milliseconds
• Number of steps observed (for statistical confidence)

Validation Benchmarks - we collect:

• Model recipe and hardware profile
• Resolution bucket and inference steps per image
• Precision and quantization level
• Average time per validation prompt in milliseconds
• Number of prompts observed

What benchmark records do NOT contain:

• No user ID, organization ID, or job ID
• No training prompts, captions, or dataset content
• No checkpoint outputs or model weights
• No API keys, tokens, or credentials
• No IP addresses or device identifiers

Benchmark data is aggregated by configuration parameters (recipe + hardware + resolution + settings). There is no way to trace a benchmark record back to a specific user or job. This anonymized data benefits the entire community by enabling accurate runtime and cost estimates.

Content Moderation

To ensure compliance with our Terms of Service, we employ automated content moderation systems that may analyze:

• Metadata and file names of submitted training data
• Training configurations for policy compliance
• Patterns indicative of unauthorized copyrighted material

This analysis is performed to:

• Detect potential violations of intellectual property rights
• Prevent use of the Service for illegal purposes
• Protect the Service and its users from harmful content

Moderation results (pass/fail status and rejection reasons) are retained for account compliance monitoring. A high rate of rejected submissions may result in account review or suspension as described in our Terms of Service.

Usage Data

We automatically collect:

• IP addresses
• Browser type and version
• Pages visited and features used
• Timestamps of interactions

3. How We Use Your Information

We use collected information to:

• Provide and maintain the Service
• Process payments and prevent fraud
• Enforce content policies through automated moderation
• Send transactional emails (job status, receipts, account notifications)
• Provide customer support
• Improve the Service
• Comply with legal obligations

How Benchmark Data Powers Runtime Estimates

When you're about to submit a training job, we show you an estimated runtime and cost. This estimate is calculated using:

• Your training parameters: Resolution buckets, sample counts, total steps, validation settings
• Community benchmarks: Average step times from anonymized data collected across all users who trained similar configurations on similar hardware

For example, if you're training a FLUX LoRA on an RTX 4090 at 1.0 megapixel resolution with batch size 4, we look up the average step time from our benchmark database for that exact combination. We then multiply by your total steps and add estimated validation time to give you a prediction like "~38 minutes, ~$0.68".

Without community benchmark data, we'd have to rely on rough estimates based only on model type. The anonymized benchmarks let us give you much more accurate predictions, saving you from cost surprises.

4. Information Sharing

We share information only with:

• Clerk: For authentication services
• Dodo Payments: For payment processing
• Infrastructure Providers: To execute your training jobs and provide managed storage as requested
• Cloudflare R2: Object storage for managed datasets
• Service Providers: For infrastructure, monitoring, and content moderation (with appropriate data processing agreements)

We do NOT sell your personal information.

5. Data Retention

• Account data: Retained while your account is active
• Waitlist emails: Retained until early access opens or April 1, 2026, then deleted
• Job metadata: Retained for 90 days for support and debugging
• Training configs: Retained only for the upload window and job execution, then deleted when the job completes
• Training data (job staging): Retained only for the upload window and job execution, then deleted when the job completes
• Managed storage datasets: Retained until your TTL expires (default 90 days), you delete them, or we delete them for non-payment of storage fees
• Training outputs: Available for download after job completion, then deleted
• Provider API keys: Deleted immediately when job attempt ends
• Invoice records: Retained for 7 years for tax compliance
• Anonymized benchmarks: Retained indefinitely (contains no personal information)

6. Data Security

We implement industry-standard security measures:

• All data transmitted over TLS/HTTPS
• Encryption at rest for sensitive data
• Cloudflare R2 encrypts stored data at rest
• Regular security audits
• Access controls and audit logging

7. Your Rights

Depending on your location, you may have the right to:

• Access: Request a copy of your personal data
• Rectification: Correct inaccurate data
• Deletion: Delete your account and associated data
• Portability: Export your data in a machine-readable format
• Object: Object to certain processing activities

To exercise these rights, contact [email protected].

8. Account Deletion

You can delete your account at any time from the dashboard or by contacting support. Upon deletion:

• Active jobs are canceled
• Personal identifiers are removed or anonymized
• Transaction records are retained with anonymized identifiers for tax compliance

9. Cookies

We use essential cookies for:

• Authentication sessions
• Security (CSRF protection)

We do not use tracking or advertising cookies.

10. International Transfers

Your data may be processed in countries other than your own. We ensure appropriate safeguards are in place for international data transfers.

11. Children's Privacy

The Service is not intended for users under 18 years of age. We do not knowingly collect data from children.

12. Changes to This Policy

We may update this Privacy Policy from time to time. Changes will be posted on this page with an updated "Last updated" date. Material changes will be communicated via email.

13. Contact Us

For privacy-related questions or to exercise your rights:

Email: [email protected]