specification: FinOps Framework specificationVersion: '1.0' schema: https://www.finops.org/framework/ provider: Poolside providerId: poolside-ai created: '2026-07-02' modified: '2026-07-02' reconciled: false tags: - AI - Foundation Models - Enterprise - On-Prem - FinOps - Cost Management - FOCUS description: >- FinOps view of Poolside spend. Because Poolside has no public price list, the dominant cost driver is a negotiated enterprise contract covering either a Cloud VPC deployment (AWS/Azure/Google) or an on-prem/air-gapped deployment (Poolside-pre-provisioned hardware or the customer's own NVIDIA GPU cluster via Helm), plus the underlying infrastructure spend for whichever path is chosen. There is no per-token metered invoice from Poolside itself to reconcile against FOCUS the way there is for a shared multi-tenant model API. The one metered, FOCUS-friendly cost surface tied to a Poolside model is third-party: calling the open-weight Laguna XS 2.1 model through OpenRouter, which bills per token under OpenRouter's own invoice, not Poolside's. notes: >- Revisit this file if Poolside ships a public, metered API tier. Until then, cost governance is a hardware/contract-sizing exercise done directly with Poolside's sales team, and infrastructure cost tracking (GPU-hours, storage, network) for the customer-controlled deployment. sources: - https://poolside.ai/enterprise - https://docs.poolside.ai/deployment/on-prem/admin - https://openrouter.ai/poolside - https://focus.finops.org/focus-specification/v1-3/ alignedWith: framework: FinOps Foundation Framework frameworkUrl: https://www.finops.org/framework/ dataSpec: FOCUS dataSpecVersion: '1.3' dataSpecUrl: https://focus.finops.org/focus-specification/v1-3/ publisherName: Poolside serviceCategory: AI and Machine Learning billingModel: pricingCategory: Contract-Based billingFrequency: Custom billingCurrency: USD chargeCategories: - Purchase - Usage meters: - name: enterprise_contract description: Negotiated annual/multi-year enterprise agreement for a Poolside deployment (Cloud VPC or on-prem/air-gapped). No published unit price. unit: contract aggregation: sum dimensions: - account - deployment - name: deployment_infrastructure description: Underlying GPU compute, storage, and network cost of running the customer-controlled Poolside deployment (cloud instances or owned hardware). unit: hours aggregation: sum dimensions: - deployment - provider - name: openrouter_laguna_tokens description: Tokens billed by OpenRouter (third party) when calling Poolside's open-weight Laguna XS 2.1 model through OpenRouter's API rather than a Poolside deployment. unit: tokens aggregation: sum dimensions: - account - model principles: - name: Visibility description: Track the enterprise contract terms and the separate infrastructure bill (cloud VPC spend or owned GPU hardware amortization) for the Poolside deployment. - name: Allocation description: Attribute deployment infrastructure cost to the team(s) or environment(s) the Poolside deployment serves, since Poolside itself does not meter per-team usage centrally. - name: Optimization description: Right-size GPU capacity against the "hundreds of developers per node" reference point Poolside publishes, and evaluate OpenRouter's metered Laguna XS 2.1 access for lightweight/evaluation workloads before committing to a full deployment. - name: Accountability description: Assign a named contract owner for the Poolside enterprise agreement and a named infrastructure owner for the deployment, since no self-serve billing dashboard exists. maintainers: - FN: Kin Lane email: kin@apievangelist.com