Vendors · Comparison
Hugging Face Inference Endpoints vs Together AI
Only Hugging Face Endpoints documents GDPR DPA; only Together AI documents HIPAA BAA and ISO 27001. Pricing, compliance and residency compared, with sources.
Last verified (older of the two profiles) How we verify
From the data
Key differences
- Pricing Hugging Face Inference Endpoints: Per instance-hour (CPU, GPU, and accelerator instances); dedicated inference from $0.033/hour (published pricing). Together AI: Per 1M tokens serverless; per GPU-hour for dedicated endpoints and clusters (published pricing).
- Category Hugging Face Inference Endpoints is a Managed self-host offering; Together AI is Dedicated GPU.
- Compliance Both document SOC 2 Type II. Only Hugging Face Inference Endpoints documents GDPR DPA. Only Together AI documents HIPAA BAA and ISO 27001.
- Data residency Hugging Face Inference Endpoints's data residency is not verified; Together AI states data residency on its own pages (not contractual).
- Jurisdiction Both parent companies are under United States jurisdiction.
Generated from the verified vendor data below; sources and dates on each vendor profile.
Hugging Face Inference Endpoints
Deploys Hub models on managed instances across AWS, Azure, and GCP with public, protected, or private (intra-region AWS/Azure PrivateLink) access. Vendor docs state SOC 2 Type 2 certification and a GDPR DPA via Enterprise Hub subscription, and that payloads are not stored (logs kept 30 days). Privacy policy: data may be stored in the US.
Together AI
Dedicated endpoints are explicitly single-tenant GPU instances with public per-hour pricing. The dedicated endpoints page says customers choose the deployment region for data residency; the terms make no geographic commitment, so residency is Stated. ISO 27001:2022 certified by A-LIGN, an ANAB-accredited body. Reserved dedicated capacity is also sold in throughput units (PTUs).
Side by side
Capabilities compared
| Hugging Face Inference Endpoints | Together AI | |
|---|---|---|
| Pricing model | Per instance-hour (CPU, GPU, and accelerator instances); dedicated inference from $0.033/hour | Per 1M tokens serverless; per GPU-hour for dedicated endpoints and clusters |
| Public pricing | Yes | Yes |
| Category | Managed self-host | Dedicated GPU |
| Deployment models | Dedicated endpointsPublic endpointProtected endpointPrivate endpoint (PrivateLink) | ServerlessDedicated endpointGPU clusters |
| Regions | United States, EU/EEA | Not verified |
| Compliance | SOC 2 Type II (stated)GDPR DPA (stated) | SOC 2 Type II (stated)HIPAA BAA (stated)ISO 27001 |
| Data residency evidence | Not verified | Stated |
| Parent jurisdiction | United States | United States |
| Sovereign for | Not recorded | Not recorded |
| Analyst note | Deploys Hub models on managed instances across AWS, Azure, and GCP with public, protected, or private (intra-region AWS/Azure PrivateLink) access. Vendor docs state SOC 2 Type 2 certification and a GDPR DPA via Enterprise Hub subscription, and that payloads are not stored (logs kept 30 days). Privacy policy: data may be stored in the US. | Dedicated endpoints are explicitly single-tenant GPU instances with public per-hour pricing. The dedicated endpoints page says customers choose the deployment region for data residency; the terms make no geographic commitment, so residency is Stated. ISO 27001:2022 certified by A-LIGN, an ANAB-accredited body. Reserved dedicated capacity is also sold in throughput units (PTUs). |
Where they diverge
Deployment differentiation
Only Hugging Face Inference Endpoints
Both
No overlap in deployment models.
Only Together AI
Read the full profiles