You own your data. We never train on it.
Use Canadian-resident or global AI routes by selecting the exact model key, with route and custody evidence on every request.
Route to the world's best models
Northern Inference processes request content in flight to route it, but does not train on customer content or keep prompts and responses by default. We keep operational metadata such as the resolved model, token counts, cost, and jurisdiction. Provider training, logging, and retention terms remain route-specific and are published on each model card.
Your prompts and completions belong to you. We route them. We do not claim them.
Northern Inference does not train models on your content. Provider data-use terms are published for each exact route.
We do not persist request content by default. It passes through our infrastructure in memory only and is gone when the request ends.
Tier 3 routes keep the NI-controlled path and selected provider region in Canada. The portal stores a tamper-evident custody record whose canonical digest is signed with NI's published Ed25519 key.
Every request carries a chain-of-custody record identifying exactly which entity, region, and jurisdiction we routed it to at each hop. Proof your security team can audit, not just a promise. Our residency guard fails closed: a Tier 3 request that cannot stay in a Canadian region is blocked, not silently rerouted. We attest the route up to the provider boundary. Inside the provider, where inference runs and how your data is handled rests on the provider's own commitments, and we surface those: every model publishes a data-use posture that cites the provider's published policy, dated and quoted verbatim, with a link to the live source. See each model's data-use posture.
At Tier 3 we route to Canadian regions such as AWS ca-central-1 and Azure Canada East. The routing layer fails closed when a request cannot be served in the selected jurisdiction. Up to the provider boundary the route is ours to attest; inside the provider we cite the provider's dated documentation on each model's data-use posture.
Every response identifies the resolved route, provider region, jurisdiction, and custody path. It also shows whether the provider call used NI-managed credentials or the customer's BYOK credential. The portal preserves the corresponding signed custody record.
For every model route we publish where training, logging, and retention stand. Each claim is cited to the provider's own dated documentation, with a link to the live source and the relevant section highlighted. Claims you can check, not assertions you can't. Browse the data-use posture.
Select your privacy level by choosing the exact model route. Customer-hosted and NI sovereign hardware are planned tiers; Canadian cloud and provider-default routes are live today.
Opt-in automatic PII detection and substitution replaces supported identifiers before the provider call, then restores them in responses. Operational audit metadata records that substitution occurred without storing the original prompt.
Runs on AWS Canada infrastructure that holds SOC 2 Type II, ISO 27001, and PCI-DSS (AWS attestations, not NI's). ITSG-33 PBMM control mapping in progress. Full status and roadmap on our Trust page.
Transparent per-token rates and clear invoicing in CAD. Credit purchases carry the disclosed 9% service fee. Usage, fees, and balance deductions remain separately reviewable.
OpenAI-compatible chat completions and Anthropic-compatible Messages endpoints support common SDKs and tools. Feature compatibility depends on the endpoint and selected provider route.
Priority access and dedicated onboarding for government and public sector organizations.
Northern Inference uses the industry-standard chat completions API format. Use an NI API key, point any compatible SDK to our endpoint, then choose an exact model route from the live catalog.
Choose the model route for every API call. Planned customer-hosted and sovereign-hardware routes, live Canadian cloud, and live provider-default access each carry an explicit tier in the route key.
Your hardware, your premises, open-source models. NI tunnels API requests into your machine; only you ever see prompts and completions. The strongest privacy guarantee we offer. Phase 5 of our roadmap.
True sovereignty, beyond residency: NI's own bare-metal in a Canadian-owned data centre. Single-tenant, we hold the keys, no US-owned cloud anywhere in the path. NI can see plaintext to route your request; the guarantee is jurisdiction and control, not cryptographic isolation. This is where sovereign AI is heading, and it is Phase 5 of our roadmap.
AWS Bedrock and Azure OpenAI routes in Canadian regions. Provider training, logging, and retention posture is published per route from the provider's current documentation.
Broader model access through NI-routed upstream providers outside the Canadian-residency boundary. This can include direct APIs, Bedrock in US regions, Azure GlobalStandard, Vertex US regions, and other provider-default routes.
Tier 3: Canadian cloud. AWS Bedrock or Azure OpenAI in ca-central-1. Provider cannot train on your data.
Residency answers where your data sits. Sovereignty answers whose law reaches it. We are standing up Tier 2 on our own Canadian hardware: bare-metal we own, in a Canadian-owned data centre, with no US-owned cloud anywhere in the path.
Single-tenant bare-metal in a locked cage in a Canadian-owned facility. Not a VM on someone else's cloud.
The planned architecture keeps encryption keys under NI control and removes a third-party cloud provider from the inference path.
The planned Tier 2 path uses NI-owned hardware in a Canadian-owned facility, reducing reliance on foreign-owned cloud infrastructure.
The planned inference engine will isolate prompt and KV-cache state per tenant. The implementation must pass security review before this tier is marked live.
Tier 4 is access. Tier 3 is Canadian residency by contract. Tier 2 is sovereignty by control, the next rung on the ladder, and the need is not unique to Canada or to controlled-goods work. Healthcare, financial services, legal, defence, and government all depend on infrastructure that cannot answer to a foreign cloud, in Canada and beyond.
Current routes available through the NI API with transparent unit pricing. Choose by provider, tier, or data residency.
Managed Canadian cloud routes. Data residency stays in Canada when the exact route is Tier 3.
Provider-default or global routes. Use when model access matters more than Canadian residency.
Use one API for the routes in the live catalog. Switch provider, tier, and region by changing the exact model key.
Uses documented OpenAI-compatible chat completions and Anthropic-compatible Messages endpoints. Supported request features depend on the endpoint and route.
The live catalog includes enabled Claude, GPT, Gemini, Llama, Mistral, and other routes. The exact model key determines provider, tier, and region.
See the provider cost and our fee separately on every request. Thinking tokens billed at the rate shown per model. No expiring credits. No surprise overages.
Choose the exact route key on every API call. Route sensitive prompts through your own hardware (coming soon), casual queries through cloud providers.
Names, emails, and identifiers are replaced with realistic fakes before reaching the model, then restored in responses. Opt-in per API key or per request.
Live dashboard with per-model, per-team cost breakdowns. Budget alerts and hard limits built in.
The NI control plane runs in Canada. Tier 3 exact routes add Canadian provider-region enforcement and fail closed on jurisdiction drift.
On supported routes, NI can mark stable prompt prefixes for provider caching automatically. Cache behavior, retention, jurisdiction, and price remain specific to the selected provider route and are disclosed in its data-use and pricing records.
Northern Inference is developer infrastructure, not a consumer chatbot.
| Northern Inference | Venice.ai | OpenRouter | Direct APIs | |
|---|---|---|---|---|
| Per-request privacy control | ||||
| Canadian data residency (Tier 3 routes) | ||||
| Transparent pricing | ||||
| Standard API format | ||||
| Per-request custody trail | ||||
| PII substitution | ||||
| No crypto required |
See provider-rated usage cost and any NI per-request fee separately. Credit purchases carry a 9% service fee. There are no monthly minimums, credits do not expire, and thinking-token rates are shown per model when applicable.
Early access members get priority onboarding
Be among the first to use Northern Inference.
Each referral moves you closer to the front of the line.