specification: API Commons Plans specificationVersion: '0.1' schema: https://raw.githubusercontent.com/api-evangelist/interface-research/main/schema/api-commons.yml#/$defs/Plans provider: LaunchDarkly providerId: launchdarkly generated: '2026-08-27' method: searched source: https://launchdarkly.com/pricing/ http_status: 200 checked: '2026-08-27' created: '2026-05-04' modified: '2026-08-27' description: >- LaunchDarkly prices on usage, not seats — service connections (one server-side SDK connected to one environment for a month) and client-side MAU are the two meters, with AI Runs as a third for AgentControl. Seats are explicitly unlimited and free on every tier. Four tiers plus one paid add-on. supersedes_note: >- Replaces the 2026-05-04 bulk-sweep version of this file, which was method:generated and carried $12 per service connection and $10 per 1k MAU. The published page on 2026-08-27 states $10 per service connection and $8.33 per 1k client-side MAU billed yearly, and the free tier's included quotas have changed as well. The old numbers were stale. currency: USD plan_count: 4 addon_count: 1 plans: - id: launchdarkly-developer name: Developer type: freemium price: '0.00' billing_period: month tagline: One platform for features and agents. Free to start. signup: self-serve, no credit card required included: - label: Seats value: unlimited note: No per-seat pricing on any tier; billing is usage-based. - label: Feature flags value: unlimited - label: SDKs value: 30 (13 client-side, 13 server-side, 4 edge) - label: Service connections value: 5 per month - label: Client-side MAU value: 1,000 per month - label: Experimentation MAU value: 100,000 per month - label: Session replays value: 5,000 per month - label: Errors value: 5,000 per month - label: Logs value: 10,000,000 per month - label: Traces value: 10,000,000 per month - label: AI Runs value: 5,000 per month - label: Projects / environments value: 1 project, 3 environments - label: AgentControl data retention value: 14 days limits: config_delivery_latency_slo: '< 300ms' overage_behavior: Ingestion stops at 5K (hard stop, no overage billing) - id: launchdarkly-foundation name: Foundation type: paid pricing_model: pay-as-you-go trial: 14 days tagline: Ship features and agents to production. Scale both without hitting a ceiling. entries: - label: Per service connection name: service_connection type: metered metric: connection-month unit: 1 price: '10.00' billing_period: month billing_note: Billed yearly. Serverless and short-lived processes bill fractionally by connection time. - label: Per 1k client-side MAU name: client_mau type: metered metric: MAU unit: 1000 price: '8.33' billing_period: month billing_note: Billed yearly. - label: Additional AI Runs name: ai_runs type: metered metric: AI Run unit: 1000 price: '5.00' billing_period: month included: - label: Service connections value: first 5 per month included - label: AI Runs value: 5,000 per month - label: Projects and environments value: unlimited - label: Session replays / errors / logs / traces value: 5K / 5K / 10M / 10M per month, then pay as you go - label: AgentControl data retention value: 30 days limits: config_delivery_latency_slo: '< 200ms' overage_behavior: '$5 per 1,000 additional AI Runs' notes: - No platform fee and no seat fee. - >- Custom packaging offered for Kubernetes, serverless, Python and Ruby architectures, which generate a high volume of server-side connections and would otherwise be penalised by per-connection pricing. - id: launchdarkly-enterprise name: Enterprise type: custom price: contact sales tagline: Full Platform controls and security for engineering teams at scale. trial: Enterprise trial available included: - Everything in Foundation - Advanced user targeting (bulk targeting, large list-based and synced segments) - Custom roles and teams (fine-grained permissions) - Workflows, scheduling and approvals - Release automation and Release Assistant - SCIM user provisioning - Custom number of AI Runs (contracted) - AgentControl data retention 100 days limits: config_delivery_latency_sla: '< 150ms (SLA, not SLO)' overage_behavior: Contracted true-up note: >- SSO and SAML are included on EVERY plan; only SCIM provisioning is gated to Enterprise. Custom roles being Enterprise-only matters for agent safety: the fine-grained resource-specifier authorization recommended in the MCP docs is not available below this tier. - id: launchdarkly-guardian name: Guardian type: addon price: contact sales requires: Enterprise tagline: Release safety on top of Enterprise — monitoring, guardrail metrics and automatic rollback. included: - Release Monitoring - Guardrail Metrics (watched during every rollout to catch regressions) - Proactive Failure Notifications - Automatic Pause or Rollback - Advanced Observability - Exposure Insights note: >- This is the tier that automates reversal. See the reversibility block in conventions/launchdarkly-conventions.yml — without Guardian, rolling back a bad release is a human action. meters: - name: Service Connection definition: One server-side SDK connected to one environment for a month. - name: Client-side MAU definition: Each unique client-side user or device per month. - name: AI Run definition: Generated when you track an LLM request or run an online eval (LLM-as-a-judge). - name: Session replay definition: One recorded user session (a single tab instance), stored for one month. - name: Error definition: One unique error or exception, stored for one month. - name: Log / Trace definition: One log line / one span, ingested, stored and searchable for one month. api_access: note: >- The REST API is not itself a metered or gated product — it is available on every tier, including the free Developer tier, and rate limits rather than plan tier are what bound its use. Individual API RESOURCES are gated by feature: workflows, scheduled changes, approvals, custom roles and teams are Enterprise, so calls to those endpoints will fail by entitlement on lower tiers. see: rate-limits/launchdarkly-rate-limits.yml