name: Parea AI FinOps description: >- FinOps Framework FOCUS-aligned cost and usage guidance for Parea AI platform consumption. Covers subscription costs, log volume tracking, and LLM proxy pass-through costs for teams managing AI engineering budgets. specificationVersion: '1.0' url: https://www.parea.ai/#pricing provider: Parea AI currency: USD billingModel: subscription costCategories: - name: Platform Subscription description: Monthly subscription fee for access to the Parea AI platform. type: fixed plans: - plan: Free monthlyCost: 0 - plan: Team monthlyCost: 150 - plan: Enterprise monthlyCost: custom - name: Log Volume description: >- Cost allocation driven by the number of trace logs recorded per month. Exceeding plan log limits may require plan upgrade. Primary cost driver for high-volume production applications. type: consumption unit: logs tiers: - name: Free Tier limit: 3000 overageCost: plan-upgrade-required plan: Free - name: Team Tier limit: 100000 overageCost: plan-upgrade-required plan: Team - name: Enterprise Tier limit: custom overageCost: negotiated plan: Enterprise - name: LLM Provider Pass-Through description: >- When using the Parea LLM proxy gateway (POST /api/parea/v1/completion), underlying LLM provider costs (OpenAI, Anthropic, etc.) are billed directly by those providers and are not included in Parea subscription fees. Teams should budget separately for model token consumption. type: pass-through providers: - OpenAI - Anthropic - Other configured LLM providers costOptimization: - recommendation: Start on Free plan during development description: >- Use the Free plan (3k logs/month) during development and testing phases to avoid subscription costs before production workloads are stable. - recommendation: Monitor log volume against plan limits description: >- Track monthly log consumption in the Parea dashboard to anticipate plan upgrade needs before hitting the 3k (Free) or 100k (Team) cap. - recommendation: Use LLM evaluation to reduce model costs description: >- Leverage Parea's evaluation and A/B testing capabilities to identify cheaper or smaller models that meet quality thresholds, reducing downstream LLM provider token costs. - recommendation: Leverage self-hosting for large deployments description: >- Enterprise teams with very high log volumes can reduce per-log costs by using Parea's self-hosting Docker option instead of the managed cloud. - recommendation: Use datasets to reduce redundant inference description: >- Build curated test datasets from production logs to run targeted evaluations rather than exhaustive inference-heavy test runs on every prompt change. allocation: dimensions: - name: Project description: Allocate log costs by Parea project (project_uuid) to map usage to product features or teams. - name: Experiment description: Tag experiment runs to associate evaluation costs with specific model or prompt change initiatives. - name: Environment description: Distinguish production vs. staging log volumes using trace metadata to isolate operational costs.