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: Affiliated Managers Group providerId: affiliated-managers-group created: '2026-05-04' generated: '2026-08-30' method: probed modified: '2026-08-30' reconciled: false source: >- live response headers observed on https://wealth.amg.com/wp-json/amgfundsdata/v1/products/funds and https://www.amg.com/wp-json/amginc/v1/affiliates, 2026-08-30 notes: >- Supersedes the 2026-05-04 bulk-sweep entry, which recorded "AMG does not operate a public developer API". That is no longer accurate: AMG serves two undocumented, anonymously callable site-backing JSON APIs (see openapi/). What remains true is that AMG documents no rate limit for them and returns no rate-limit signal at runtime. limit_count is an honest zero — checked, nothing published. description: >- AMG publishes no rate-limit policy for either site-backing API, and no rate-limit response header was observed. An automated consumer has no runtime signal to back off on and must self-throttle. tags: - Rate Limiting - Financial Services - Asset Management sources: - https://wealth.amg.com/wp-json/amgfundsdata/v1 - https://www.amg.com/wp-json/amginc/v1 limit_count: 0 limits: [] response_headers: observed: [] absent: - RateLimit-Limit - RateLimit-Remaining - RateLimit-Reset - X-RateLimit-Limit - X-RateLimit-Remaining - X-RateLimit-Reset - Retry-After status_on_exhaustion: null responseCodes: {} evidence: - url: https://wealth.amg.com/wp-json/amgfundsdata/v1/products/funds status: 200 rate_limit_headers: none note: Response carried CloudFront cache headers (x-cache, age) and HSTS, but no rate-limit family header. - url: https://www.amg.com/wp-json/amginc/v1/affiliates status: 200 rate_limit_headers: none policies: - name: No published rate-limit policy description: >- Neither AMG site publishes limits, quotas or a fair-use statement for its JSON endpoints, and no developer terms cover programmatic use. Consumers should throttle conservatively and cache; the full products dataset is roughly 2.4 MB per call and is edge-cached, so repeated polling is both expensive and pointless.