finops: summary: > Pelias is MIT-licensed open-source software with no software licensing costs. The primary cost model is infrastructure (self-hosted) or subscription-based (Geocode Earth hosted service). Because it is powered entirely by open data, there are no data licensing fees. The hosted Geocode Earth service provides predictable monthly billing starting at $100/month. costModels: - model: Self-Hosted description: > Run Pelias on your own servers or cloud infrastructure. No software cost. Costs are driven by Elasticsearch cluster sizing, data storage for imported datasets (OpenStreetMap, OpenAddresses, etc.), compute, and bandwidth. costDrivers: - Elasticsearch cluster (dominant cost) - Server/VM compute for API, Placeholder, PIP, Interpolation services - Storage for geocoding datasets (can be several hundred GB to TB) - Bandwidth for import and serving - Operational overhead for maintenance and updates estimatedMonthlyCost: "Varies; small deployments ~$50-300/month cloud infra" - model: Geocode Earth Hosted description: > Pay-per-plan subscription to the commercial Pelias API operated by the original Pelias creators. Predictable monthly billing with no infrastructure management required. tiers: - name: Lite monthlyPrice: 100 currency: USD requestsIncluded: 150000 costPerThousandRequests: 0.67 - name: Basic monthlyPrice: 200 currency: USD requestsIncluded: 400000 costPerThousandRequests: 0.50 - name: Advanced monthlyPrice: 500 currency: USD requestsIncluded: 2000000 costPerThousandRequests: 0.25 - name: Custom monthlyPrice: negotiable currency: USD requestsIncluded: "50000000+" costPerThousandRequests: negotiable costOptimizationTips: - Cache geocoding results client-side or server-side; Geocode Earth explicitly permits storing results indefinitely. - Use autocomplete endpoint only for interactive search; batch offline geocoding against the search endpoint. - Self-host for high-volume use cases where cloud infrastructure cost is less than hosted plan cost. - Geocode Earth offers discounts for open-source projects and non-profits — contact them directly. - Use structured geocoding (/v1/search/structured) when input data is already parsed to improve match accuracy and reduce retry requests. - Import only the data sources relevant to your geography to reduce Elasticsearch index size and query costs. openSourceDiscounts: available: true notes: > Geocode Earth offers free or discounted plans for qualifying open-source projects. Contact team@pelias.io or via https://geocode.earth for details. trialAvailability: available: true duration: "2 weeks" creditCardRequired: false url: https://geocode.earth