generated: '2026-08-04' method: probed source: response headers observed on https://gunzscan.io on 2026-08-04 summary: >- Gunzilla publishes no rate-limit documentation. GUNZScan does signal quota in response headers on both its REST v2 and GraphQL surfaces, so the limits below were read off the wire rather than off a docs page. The GUNZ chain RPC node returns no rate-limit headers at all; the documentation only implies limits exist by offering IP allow-listing to remove them. rate_limits: - api: GUNZScan Explorer API surface: REST v2 (/api/v2/*) limit_count: 3000 window: rolling (server-reported reset in milliseconds) scope: per client IP signaled: true headers: limit: x-ratelimit-limit remaining: x-ratelimit-remaining reset: x-ratelimit-reset observed: x-ratelimit-limit: '3000' x-ratelimit-remaining: '2935' x-ratelimit-reset: '41304' evidence: GET https://gunzscan.io/api/v2/stats - api: GUNZScan Explorer API surface: GraphQL (/api/v1/graphql) limit_count: 500 window: rolling (server-reported reset in milliseconds) scope: per client IP signaled: true headers: limit: x-ratelimit-limit remaining: x-ratelimit-remaining reset: x-ratelimit-reset observed: x-ratelimit-limit: '500' x-ratelimit-remaining: '481' x-ratelimit-reset: '2730274' additional_control: kind: query-complexity max_complexity: 100 note: enforced per operation, independent of the request quota evidence: POST https://gunzscan.io/api/v1/graphql - api: GUNZ Chain JSON-RPC API surface: https://rpc.gunzchain.io/ext/bc/.../rpc limit_count: null signaled: false scope: per client IP (inferred) note: >- No x-ratelimit-* headers are returned. Gunzilla's documentation confirms limits exist and offers to remove them for partners. quote: >- "If you intend to utilize our RPC endpoint, please provide the IP addresses of your services so that we can whitelist them. This will eliminate potential limit hits." source: https://gunbygunz.com/documentation/ exemption_channel: contact the Gunzilla team to have service IPs allow-listed gaps: - No published rate-limit page, no documented numbers, no documented 429 behaviour, no Retry-After header observed on any surface. - The two GUNZScan surfaces carry different quotas (3000 REST vs 500 GraphQL) with no documentation saying so - a client that budgets from the REST number will be throttled six times sooner on GraphQL.