generated: '2026-09-19'
method: searched
source: https://moltrust.ch/pricing.html + https://moltrust.ch/docs/caep.html + openapi/moltrust-ch-moltguard-openapi.yml
  (429 RateLimitError) + openapi/moltrust-ch-openapi.yml (operation prose) + a2a/moltrust-ch-agent-card.json + live
  probe 2026-09-19
corroboration: Live GET https://api.moltrust.ch/identity/agent-types (200) and /skill/trust-score/{did} (200) returned
  NO RateLimit-*, X-RateLimit-* or Retry-After headers; no 429 was provoked (probing to exhaustion is out of bounds).
  The only runtime signal a client gets is the 429 body.
confidence: medium
confidence_note: Numbers come from the provider's own pages and spec prose and are internally inconsistent (60/hour
  on the pricing page vs 100/day in the agent card). Header names cannot be recorded because none are published
  or observed.
limit_count: 6
scopes:
- scope: per-key (free tier)
  surface: MolTrust API
  window: 1 hour
  limit: 60
  unit: requests
  burst: null
  status_on_exhaustion: 429
  headers: []
  retry_after: null
  source: https://moltrust.ch/pricing.html ("60 free calls/hour once your DID is registered")
  conflict: agent card securitySchemes.apiKey.description says "Free tier is 100 requests/day"
- scope: per-account (credits)
  surface: MolTrust API
  window: monthly floor / per call
  limit: 30 credits/month floor; 100 credits at registration; 0-2 credits per call
  unit: credits
  status_on_exhaustion: not documented (likely 402/403 - unverified)
  headers: []
  source: https://moltrust.ch/pricing.html + GET /credits/pricing
  note: Quota-shaped rather than rate-shaped; balance readable at GET /credits/balance/{did} (free).
- scope: per-address (free MoltGuard reads)
  surface: MoltGuard /guard/api/agent/score-free/{address}, /guard/api/market/check-free/{marketId}
  window: 10 minutes
  limit: 1
  unit: requests
  status_on_exhaustion: 429
  body: 'RateLimitError {"message":"Free tier: max 1 request(s) per 10 minutes. Use x402 paid endpoints for unlimited
    access."}'
  headers: []
  retry_after: null
  source: openapi/moltrust-ch-moltguard-openapi.yml#getAgentScoreFree + /guard/api/info
  method: derived
- scope: per-IP
  surface: MoltGuard GET /guard/skill/audit
  window: 1 hour
  limit: 5
  unit: requests
  status_on_exhaustion: 429
  headers: []
  source: https://moltrust.ch/llms.txt ("5 per hour per IP") + /guard/skill/info + spec 429
- scope: per-DID
  surface: MolTrust API GET /caep/pending/{did}
  window: 1 hour
  limit: 120
  unit: requests
  status_on_exhaustion: 429 (undeclared in spec)
  headers: []
  source: https://moltrust.ch/docs/caep.html ("Rate-limit 120/h per DID") + operation description
- scope: per-IP
  surface: MoltGuard POST /guard/hackathon/register
  window: null
  limit: null
  unit: null
  status_on_exhaustion: 429
  body: '"IP rate limit exceeded"'
  headers: []
  source: openapi/moltrust-ch-moltguard-openapi.yml#registerHackathon
validated_and_rate_limited_reads:
  operations:
  - get_trust_score_skill_trust_score__did__get
  - get_swarm_graph_swarm_graph__did__get
  source: 'spec descriptions: "Validated and rate limited: the handler writes trust_score_cache rows, so an unbounded
    stream of junk DIDs was both a compute cost and a way to grow the cache table."'
  limit_published: false
paid_rail:
  note: x402-paid MoltGuard endpoints are documented as "unlimited access" (RateLimitError message) - the limit
    is economic (USDC per call), not a window.
mcp_surface:
  endpoint: https://api.moltrust.ch/mcp
  limits_published: false
  note: tools/list answered anonymously with no limit headers; tool calls inherit the REST limits of the key they
    carry.
terms: Terms 4 "No Circumventing Rate Limits" prohibits bypassing limits and quotas.