Everything in ARK Core plus the managed platform services enterprises expect — identity, logging, telemetry, monitoring, and a built-in Portal — deployed on your infrastructure, configured to your standards, and handed over to your team to operate. Software Support SLA on ARK components; your infrastructure, your operational control.
ARK Tailored is the primary offering for organisations that need the full ARK Platform — runtime plus managed services — running inside their own perimeter. It fits three patterns.
Full platform inside your VPC or on-premises, integrated with your IAM, logging, and monitoring stack. No data leaves your perimeter.
Use ARK as the inference backbone for a multi-tenant SaaS offering. OpenAI v1 / Anthropic-compatible API, session-level isolation, predictable density economics.
Sovereign/regional cloud providers deploy ARK on their infrastructure and resell managed inference to regulated-sector clients.
ARK Tailored ships with the full set of platform service components pre-configured for your environment. Already running your own Keycloak, ELK, or Prometheus/Grafana? Swap any platform service with your own tooling — ARK Tailored integrates into your existing stack rather than replacing it.
ARK Runtime — the inference engine with heterogeneous GPU support, elastic hot-scaling, and session-level KV isolation
Compute Nodes — execution runtimes bound to specific devices
Supervisor Node — orchestration with master-master replication
API Gateway — edge proxy + OpenAI v1 / Anthropic-compatible API service
Model Storage — Hugging Face-format shared repository (use ARK's or point at your own backend)
Identity Service (Keycloak) — tenancy, realms, API keys, credit balances
Logging Service (ELK) — centralised log collection
Telemetry Service (Prometheus + Grafana) — infra and service metrics, nvidia-smi integration
Monitoring Service — KPI thresholds, alerts, webhooks
Portal Service — built-in chat, image, and embedding workspace for non-technical teams
ARK Tailored deploys inside your perimeter: full platform in the middle, your applications and business systems on the outside, and nothing crosses the boundary. Everything that would normally call a public LLM API calls ARK instead.
ARK Tailored is designed for organisations that want the full platform on their own terms. ARK configures every component to your environment and gets it production-ready — then hands over to your team for day-to-day operations, with a Software Support SLA covering ARK components.
ARK deploys the full platform on your cloud or on-prem infrastructure, configured to your identity, logging, telemetry, and networking standards.
We ensure every component is operational, models are loaded, benchmarks pass, and your team has the runbooks and training to take it over.
Your team operates the platform day-to-day. ARK provides a Software Support SLA on ARK components, with response times scaled to your GPU license volume.
Need ARK to run it for you instead? Add managed services on top and we'll handle ongoing operations, model updates, and capacity planning.
ARK Tailored is sold on a monthly or annual platform license, sized by GPUs under license. Add-on modalities and extended model catalogues are priced per GPU per month. Contact sales for a quote scoped to your deployment.
Priced per GPU under license. Base allocation includes 10 models and text modality.
Software Support SLA on ARK components. Response times and escalation paths scale with deployment size.
Upgrade from 10-model base to a 10–20 model catalogue.
Per additional modality (image, vision, embeddings). Text is always included.
One-time installation guidance packages and optional managed services available on request.
Choose ARK Tailored when you want the full managed platform on your infrastructure. Choose ARK Core when you have a mature platform team and only need the runtime. Choose ARK Cloud for zero-ops serverless inference.
Per-token pricing. Best-effort 99% availability. Zero ops.
Annual license. Integrate with your own platform stack.
Runtime + platform services, deployed and supported by ARK.
Tell us about your environment, GPU footprint, and integration requirements. We'll come back with a deployment plan, timeline, and quote.