Together AI

Open-source LLM inference and fine-tuning platform for enterprise AI builders

Modal

Serverless GPU compute platform for AI model training, inference, and data pipelines

Together AI

Caution

10/25

Modal

Caution

9/25

Score Breakdown

DimensionTogether AIModal

Data Residency

Where is your data stored and processed?

Together AI: All inference and fine-tuning workloads processed in US data centres. No EU data residency option. GDPR transfer mechanisms (SCCs) required for European personal data processing.

Modal: US-only infrastructure. No EU data centre options. Not suitable for GDPR personal data processing without SCCs and TIA.

1/5

All inference and fine-tuning workloads processed in US data centres. No EU data residency option. GDPR transfer mechanisms (SCCs) required for European personal data processing.

1/5

US-only infrastructure. No EU data centre options. Not suitable for GDPR personal data processing without SCCs and TIA.

Legal Jurisdiction

Which laws govern the company and your data?

Together AI: Incorporated in Delaware, US. Subject to CLOUD Act. GDPR DPA available for enterprise customers contractually, but US jurisdiction governs. No EU subsidiary structure.

Modal: Delaware incorporation, US jurisdiction, CLOUD Act applies. Basic GDPR privacy documentation. No EU subsidiary.

2/5

Incorporated in Delaware, US. Subject to CLOUD Act. GDPR DPA available for enterprise customers contractually, but US jurisdiction governs. No EU subsidiary structure.

2/5

Delaware incorporation, US jurisdiction, CLOUD Act applies. Basic GDPR privacy documentation. No EU subsidiary.

Data Retention & Training

Is your data used for model training?

Together AI: Clear policy that customer data and fine-tuning datasets are not used to train shared models. Fine-tuned model weights are customer property. Good baseline for data isolation, but residency limitation remains.

Modal: Customer compute workloads are isolated and not used for shared training. Workload outputs are customer-controlled. Data governance documentation is lighter than enterprise-focused providers.

4/5

Clear policy that customer data and fine-tuning datasets are not used to train shared models. Fine-tuned model weights are customer property. Good baseline for data isolation, but residency limitation remains.

3/5

Customer compute workloads are isolated and not used for shared training. Workload outputs are customer-controlled. Data governance documentation is lighter than enterprise-focused providers.

Certifications

ISO 27001, SOC 2, Cyber Essentials, etc.

Together AI: No published ISO 27001, SOC 2 Type II, or independent security certifications as of early 2026. Security practices are self-attested. Significant gap for enterprise procurement in regulated industries.

Modal: No published independent certifications. Developer-focused startup with self-attested security practices. Not yet suitable for regulated industry enterprise procurement.

1/5

No published ISO 27001, SOC 2 Type II, or independent security certifications as of early 2026. Security practices are self-attested. Significant gap for enterprise procurement in regulated industries.

1/5

No published independent certifications. Developer-focused startup with self-attested security practices. Not yet suitable for regulated industry enterprise procurement.

Regulatory Fit

Suitability for regulated industries and professional services

Together AI: Not suitable for GDPR-regulated personal data processing without SCCs and TIA. Good for development, prototyping, and non-personal-data AI workloads. US jurisdiction and absence of certifications limit suitability for regulated EU industries.

Modal: Suited to AI developers and startups building non-personal-data workloads. US jurisdiction and absence of certifications limit suitability for regulated EU industries. Good for R&D and prototyping contexts.

2/5

Not suitable for GDPR-regulated personal data processing without SCCs and TIA. Good for development, prototyping, and non-personal-data AI workloads. US jurisdiction and absence of certifications limit suitability for regulated EU industries.

2/5

Suited to AI developers and startups building non-personal-data workloads. US jurisdiction and absence of certifications limit suitability for regulated EU industries. Good for R&D and prototyping contexts.

Total Score

10/25

9/25

Best For

Together AI

Best for teams on a tight budget.

Modal

Best for privacy-conscious teams who need strong data retention controls; teams on a tight budget.

Detailed Comparison

Modal vs Together AI: Trust & Compliance Comparison

Modal (Modal Labs, US) scores 9/25 overall with a Review Required (Caution) trust badge. Serverless GPU compute platform for AI model training, inference, and data pipelines. Together AI (Together AI, US) scores 10/25 with a Review Required (Caution) trust badge. Open-source LLM inference and fine-tuning platform for enterprise AI builders.

Dimension-by-Dimension Breakdown

#### Data Residency

Both score equally at 1/5.

●Modal (1/5): US-only infrastructure. No EU data centre options. Not suitable for GDPR personal data processing without SCCs and TIA.

●Together AI (1/5): All inference and fine-tuning workloads processed in US data centres. No EU data residency option. GDPR transfer mechanisms (SCCs) required for European personal data processing.

#### Legal Jurisdiction

Both score equally at 2/5.

●Modal (2/5): Delaware incorporation, US jurisdiction, CLOUD Act applies. Basic GDPR privacy documentation. No EU subsidiary.

●Together AI (2/5): Incorporated in Delaware, US. Subject to CLOUD Act. GDPR DPA available for enterprise customers contractually, but US jurisdiction governs. No EU subsidiary structure.

#### Data Retention & Training

Together AI leads with 4/5 vs 3/5.

●Modal (3/5): Customer compute workloads are isolated and not used for shared training. Workload outputs are customer-controlled. Data governance documentation is lighter than enterprise-focused providers.

●Together AI (4/5): Clear policy that customer data and fine-tuning datasets are not used to train shared models. Fine-tuned model weights are customer property. Good baseline for data isolation, but residency limitation remains.

#### Certifications

Both score equally at 1/5.

●Modal (1/5): No published independent certifications. Developer-focused startup with self-attested security practices. Not yet suitable for regulated industry enterprise procurement.

●Together AI (1/5): No published ISO 27001, SOC 2 Type II, or independent security certifications as of early 2026. Security practices are self-attested. Significant gap for enterprise procurement in regulated industries.

#### Regulatory Fit

Both score equally at 2/5.

●Modal (2/5): Suited to AI developers and startups building non-personal-data workloads. US jurisdiction and absence of certifications limit suitability for regulated EU industries. Good for R&D and prototyping contexts.

●Together AI (2/5): Not suitable for GDPR-regulated personal data processing without SCCs and TIA. Good for development, prototyping, and non-personal-data AI workloads. US jurisdiction and absence of certifications limit suitability for regulated EU industries.

Overall Verdict

Modal and Together AI are closely matched on trust and compliance, with scores of 9/25 and 10/25 respectively. The right choice depends on your specific regulatory requirements and existing technology stack.

Frequently Asked Questions

Which is better for EU compliance, Together AI or Modal?

Together AI has a TrustKit score of 10/25 while Modal scores 9/25. Together AI currently rates higher across data residency, legal jurisdiction, data retention, certifications, and regulatory fit.

How do Together AI and Modal compare on data residency?

Together AI scores 1/5 for data residency (All inference and fine-tuning workloads processed in US data centres. No EU data residency option. GDPR transfer mechanisms (SCCs) required for European personal data processing.), while Modal scores 1/5 (US-only infrastructure. No EU data centre options. Not suitable for GDPR personal data processing without SCCs and TIA.).

Are Together AI and Modal GDPR compliant?

Both tools are assessed across five compliance dimensions. Together AI has a regulatory fit score of 2/5 and Modal scores 2/5. Check the full comparison above for a detailed breakdown.

Explore Each Tool