Vendors · Comparison
Nebius vs Together AI
Nebius's parent is in the Netherlands; Together AI's is in the United States. Pricing, compliance, regions and residency compared, every fact sourced and dated.
Last verified (older of the two profiles) How we verify
Analysis
What separates them
Both sell single-tenant inference, but they sit under different law. Nebius is EU-parented: its services agreement contracts non-US, non-Israeli customers through Nebius B.V. under Dutch law. Together AI is operated by Together Computer, Inc., a Delaware corporation, and its terms are governed by California law.
That Dutch contract does not make Nebius EU sovereign. US customers who registered for Token Factory on or after 15 September 2026 contract with Nebius Inc. under the same agreement, and Nebius makes no sovereignty claim on its enterprise inference page. This site records Nebius as EU-parented, not as sovereign for the EU.
The isolation story is similar. Nebius offers shared or single-tenant endpoints; Together says every dedicated deployment runs on single-tenant, isolated GPUs.
Residency is where they part. Section 6.1 of the Nebius Data Processing Agreement commits dedicated endpoints to the customer-selected region, which is Contractual on our evidence levels. Together’s dedicated page says customers choose the deployment region for data residency, but its terms make no geographic commitment and no regions are named, so it is Stated.
Pricing differs in shape too. Together publishes per-GPU-hour rates and sells reserved capacity in throughput units. Nebius bills dedicated endpoints while a replica runs; its rates are not verified here.
Last verified (every source below) How we verify
Sources (7)
- https://docs.nebius.com/legal/agreement (checked 2026-10-22)
- https://www.together.ai/terms-of-service (checked 2026-10-22)
- https://nebius.com/services/token-factory/enterprise-grade-inference (checked 2026-10-22)
- https://www.together.ai/dedicated-endpoints (checked 2026-10-22)
- https://docs.nebius.com/legal/dpa (checked 2026-10-22)
- https://www.together.ai/pricing (checked 2026-10-22)
- https://docs.tokenfactory.nebius.com/ai-models-inference/dedicated-endpoints/billing-policy (checked 2026-10-22)
From the data
Key differences
- Pricing Nebius: Per 1M input/output tokens; dedicated endpoints billable while a replica runs (unit not published) (publication not verified). Together AI: Per 1M tokens serverless; per GPU-hour for dedicated endpoints and clusters (published pricing).
- Category Both are Dedicated GPU offerings.
- Compliance Both document SOC 2 Type II, ISO 27001 and HIPAA BAA. Only Nebius documents SOC 3 and GDPR DPA.
- Data residency Nebius commits to data residency in its contract terms; Together AI states data residency on its own pages (not contractual).
- Jurisdiction Parent jurisdiction: Nebius Netherlands, Together AI United States.
Generated from the verified vendor data below; sources and dates on each vendor profile.
Nebius
Token Factory offers shared or single-tenant endpoints with optional regional deployment in EU or US data centers (Finland, France, US). Non-US customers contract with Nebius B.V. under Dutch law; US Token Factory customers registering from 15 Sep 2026 contract with Nebius Inc. PHI use requires a signed BAA (HIPAA guideline). Per-token prices require sign-in.
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
| Nebius | Together AI | |
|---|---|---|
| Pricing model | Per 1M input/output tokens; dedicated endpoints billable while a replica runs (unit not published) | Per 1M tokens serverless; per GPU-hour for dedicated endpoints and clusters |
| Public pricing | Not verified | Yes |
| Category | Dedicated GPU | Dedicated GPU |
| Deployment models | Serverless APIDedicated endpoints | ServerlessDedicated endpointGPU clusters |
| Regions | EU/EEA, United States | Not verified |
| Compliance | SOC 2 Type II (stated)SOC 3 (stated)ISO 27001 (stated)GDPR DPAHIPAA BAA | SOC 2 Type II (stated)HIPAA BAA (stated)ISO 27001 |
| Data residency evidence | Contractual | Stated |
| Parent jurisdiction | Netherlands | United States |
| Sovereign for | Not recorded | Not recorded |
| Analyst note | Token Factory offers shared or single-tenant endpoints with optional regional deployment in EU or US data centers (Finland, France, US). Non-US customers contract with Nebius B.V. under Dutch law; US Token Factory customers registering from 15 Sep 2026 contract with Nebius Inc. PHI use requires a signed BAA (HIPAA guideline). Per-token prices require sign-in. | 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 Nebius
Both
No overlap in deployment models.
Only Together AI
Read the full profiles