// ============================================================ // Attic Standard MCP Server — Brand and product constants // ============================================================ // One place for every name, link and price the server prints. // Technical identifiers (repo, npm package, Railway URL and the // _atom_api_key field) keep their original names so existing // connections and directory listings keep working. // ============================================================ export const SERVER_NAME = "attic-standard-mcp"; export const SERVER_VERSION = "2.1.0"; export const BRAND = "Attic Standard"; export const PRO_NAME = "Attic Standard MCP PRO"; export const PRO_PRICE = "$500/month"; export const SITE = "https://atticstandard.com"; export const MCP_PAGE = `${SITE}/mcp`; export const METHODOLOGY_PAGE = `${SITE}/methodology`; /** Public icon served by this server (assets/icon.png). */ export const ICON_PATH = "/favicon.png"; export const PUBLIC_URL = "https://mcp.atticstandard.com"; export const UPGRADE_LABEL = `[${PRO_NAME}]`; export const UPGRADE_MESSAGE = `Vendor names, model names and SKU-level prices are available in ${PRO_NAME} (${PRO_PRICE}). ${MCP_PAGE}`; export const INDEX_FAMILIES = [ "Modality", "Channel", "Tier", "License", "Origin", "Use case", ] as const; /** Readable names for the channel codes stored in vendor_registry.vendor_type. */ export const CHANNEL_LABELS: Record = { DEV: "Model developer", CLD: "Cloud marketplace", PLT: "Inference platform", NCL: "Neocloud", }; /** Published tier names for the tier codes stored in model_registry.tier. */ export const TIER_LABELS: Record = { FLG: "Flagship", FTR: "Flagship", COR: "Core", MID: "Core", CMP: "Compact", BDG: "Compact", }; /** Published names for model_registry.license_class. */ export const LICENSE_LABELS: Record = { open: "Open weights, permissive license", "open-weight": "Open weights, restricted license", "non-commercial": "Non-commercial license", proprietary: "Proprietary", }; /** Market KPI definitions, keyed by kpi_market_snapshot.kpi_code (methodology section 4.2). */ export const MARKET_KPIS: Record = { output_premium: { label: "Output premium", group: "Price structure", definition: "The share of token models whose output rate is more than four times their input rate." }, caching_discount: { label: "Caching discount", group: "Price structure", definition: "The median saving of cached input against the same model's standard input rate, across models sold with both." }, caching_availability: { label: "Caching availability", group: "Price structure", definition: "The share of token models sold with a cached input rate by at least one company." }, repricing_activity: { label: "Repricing activity", group: "Price dynamics", definition: "The share of listings present now and thirteen weeks earlier whose price made a lasting change in between, with the share of those changes that were cuts." }, repricing_depth: { label: "Repricing depth", group: "Price dynamics", definition: "The median size of the lasting price changes made over the last thirteen weeks." }, post_launch_drift: { label: "Post-launch drift", group: "Price dynamics", definition: "The share of models launched four to thirteen weeks ago whose market price is now below their launch week." }, multivendor_spread: { label: "Multi-vendor spread", group: "Competition", definition: "For models sold by two or more companies, the median ratio of the dearest company's input price to the cheapest's." }, first_party_premium: { label: "First-party premium", group: "Competition", definition: "The share of models priced higher by their creator than by the median independent host, an inference platform or neocloud selling the same model." }, marketplace_premium: { label: "Marketplace premium", group: "Competition", definition: "The share of models priced higher on a cloud marketplace than by the median independent host selling the same model." }, }; export const BASE_MONTH = "2026-05"; export const BASE_MONTH_END = "2026-06-01"; export const BENCHMARK_NOTE = "Benchmark: the chained level, expressed as May 2026 = 100 (the average of that month's weeks); it moves only when models present in consecutive weeks are repriced, and each model joins from its second priced week. " + "Spot: what the market charges this week, in dollars; each model is taken at the median of its vendors' prices, and the spot is the median across those models, with the 25th and 75th percentiles. " + "Spot changes are measured against the spot's own May 2026 average, four published weeks back and one week back, so they reflect models arriving and leaving as well as repricing. " + "Where a basket averaged fewer than 15 models in May 2026, or today's count differs from that average by more than half, changed_since_base is true: read spot changes with that in view.";