Inference marketplace ยท Private alpha
Paid plans include a practical inference allowance for evaluation and everyday workloads. Choose TEKIZ.AI-hosted capacity, connect Ollama or another approved provider, or scope a private pilot around your model, processing location, and capacity needs.
Pilot availability and pricing are confirmed before you commit. Explore the marketplace.
Connection plans
Every paid plan includes TEKIZ.AI-hosted inference when live capacity is available. You can also connect the model services and local models you already use. TEKIZ.AI manages routing rules, access keys, and request records. A connected provider bills its own model usage. Your selected provider processes your requests under its data-handling terms.
Checking availability. Status
Use your own providers
Run a local model directly for a single-model workflow. Use TEKIZ.AI when your team needs one API for approved local and cloud services, with keys, routing rules, and a request record.
Lite, Pro, and Max include hosted inference within their monthly request allowances when live capacity is available. Usage through a provider account you connect stays on that provider's bill. Need a reviewed private setup and one bill? Request managed access. It is a separate private pilot.
For individuals and small businesses who bring their own model accounts or local models.
Shared routing and visibility for professionals and small teams running a real workflow.
Higher-volume routing with direct help for private setup and rollout.
What your plan covers
Use different approved services for each customer record, case file, or financial workflow. TEKIZ.AI applies the rules your team sets for that work.
A provider change, outage, or policy decision can affect every workflow at once.
Your team chooses the services that are allowed for the work. If no approved choice is available, the request can stop for a person to decide.
Your provider account, provider bill, approval rules, and the record of what happened remain under your control.
requests rely on one provider account
requests remain on the first choice; 6,000 can use 2 other approved choices
Illustrative only. Actual routing depends on your rules, task requirements, connected accounts, service health, and provider limits. TEKIZ.AI does not increase a provider's quota.
Business plans
Legal, healthcare, financial, and public-sector teams need clear answers about where work may go, who approved it, and what they can show when someone asks.
Business governance is priced around seats and accountability. Request volume remains part of capacity planning. A Premium sovereign deployment is offered only after its Canadian residency, encryption, and evidence controls are proven for that deployment.
Managed model access
Managed access is separate from the routing plans above. Tell us which models and usage you need; we will confirm availability and pricing before you pay.