Docs/Inference

Inference

Nexlayer Inference is a fully-managed training & inference platform. Bring your model and your workload; Nexlayer builds the stack around it and runs it in production.

  • Your modelsOpen-weight models or your own fine-tunes.
  • Training and servingTrain, fine-tune and serve in one managed platform.
  • Standard endpointsOpenAI-compatible and gRPC, so your app code doesn't change.
  • Sized to your workloadCapacity comes from a benchmark of your real traffic and quality bar.
  • Single-tenant, in the USReserved for you alone, with private networking and retention controls.
  • Next to your appThe app that calls the model runs on Nexlayer too, operated together.

Run your model on Nexlayer

Inference is an Enterprise product. Tell us about your workload.

Go to Nexlayer Inference

Calling a hosted model API from your app works on every plan: add the key as a secret and deploy as usual. Quickstart