specification: API Commons Rate Limits specificationVersion: '0.1' schema: https://raw.githubusercontent.com/api-evangelist/interface-research/main/schema/api-commons.yml#/$defs/RateLimits provider: Google Data Studio providerId: google-data-studio created: '2026-05-04' # Provenance stamped 2026-08-11: this artifact was written by the API Evangelist # bulk sweep dated 2026-05-04, not harvested from the provider. See roadmap#35. method: generated modified: '2026-05-05' reconciled: true tags: - Business Intelligence - Reporting - Dashboards - Google Cloud - Rate Limiting description: Looker Studio (Data Studio) UI rate limits are enforced server-side and not numerically published; query throughput is principally bounded by the underlying data source. Looker Studio APIs and Looker (the platform) ship with monthly Query API and Admin API call allocations that vary by edition. Per-edition allocations are listed below. sources: - https://cloud.google.com/looker/pricing - https://developers.google.com/looker-studio/integrate - https://cloud.google.com/looker/docs/admin-panel-database-connection-pooling responseCodes: throttled: 429 quotaExceeded: 403 limits: - name: Looker Query API - Standard edition scope: instance metric: requests_per_month limit: 1000 timeFrame: month - name: Looker Admin API - Standard edition scope: instance metric: requests_per_month limit: 1000 timeFrame: month - name: Looker Query API - Enterprise edition scope: instance metric: requests_per_month limit: 100000 timeFrame: month - name: Looker Admin API - Enterprise edition scope: instance metric: requests_per_month limit: 10000 timeFrame: month - name: Looker Query API - Embed edition scope: instance metric: requests_per_month limit: 500000 timeFrame: month - name: Looker Admin API - Embed edition scope: instance metric: requests_per_month limit: 100000 timeFrame: month - name: Conversational Analytics input tokens (Viewer) scope: user metric: tokens_per_month limit: 1000000 timeFrame: month notes: Token allocation per Viewer user per month; charged $3/M input tokens above (post-Oct 2026). - name: Conversational Analytics output tokens (Viewer) scope: user metric: tokens_per_month limit: 20000 timeFrame: month notes: Charged $20/M output tokens above (post-Oct 2026). - name: Conversational Analytics input tokens (Standard) scope: user metric: tokens_per_month limit: 2000000 timeFrame: month - name: Conversational Analytics output tokens (Standard) scope: user metric: tokens_per_month limit: 40000 timeFrame: month - name: Conversational Analytics input tokens (Developer) scope: user metric: tokens_per_month limit: 4000000 timeFrame: month - name: Conversational Analytics output tokens (Developer) scope: user metric: tokens_per_month limit: 80000 timeFrame: month - name: Looker Studio UI rate scope: user metric: requests_per_minute limit: see Looker Studio internal limits notes: Looker Studio enforces internal request smoothing; effective throughput is bounded by the connected data source's quota. policies: - name: Data-source bounded description: Looker Studio dashboards make queries against the underlying data source (BigQuery, Sheets, MySQL, etc.); the data source's own quotas are usually the binding constraint, not Looker Studio itself. - name: API monthly buckets description: Looker Query / Admin API call counts reset monthly per edition. Add-on call packs may be available via sales for higher volumes. - name: Token meter for Conversational Analytics description: Free until 2026-09-30; afterwards billed at $3/M input + $20/M output beyond per-user monthly allocations. - name: Backoff description: On 429 / 403 quotaExceeded, retry with exponential backoff; cache dashboard data via Looker's data action / extract scheduling features rather than re-querying live. maintainers: - FN: Kin Lane email: kin@apievangelist.com