Reference

Limits and quotas

Consolidated overview of every limit on the knowmind platform. Values as of 23 May 2026 — changes are tracked in the changelog.

ScopeLimitSource
Memories, Private plan100Plan matrix
Memories, Pro and aboveunlimitedPlan matrix
API rate limit, Business API plan60 requests per minute per tokenPlan matrix
API rate limit, Enterprise600 requests per minute per tokenPlan matrix
Maximum document size (upload)10 MBMemory service
Maximum memory size (store_memory)100 KBMemory service
Recall — number of hits (k)1 to 25MCP tool schema
Recall — graph hops0 to 3MCP tool schema
Chunk size on uploadabout 500 tokens with 50 tokens overlapRecall pipeline
Embedding dimension1024 (multilingual-e5-large)Recall pipeline
Webhook body size1 MBWebhook worker
Webhook timeout15 secondsWebhook worker
Webhook retries6 attempts (immediate, 1 min, 5 min, 30 min, 2 h, 12 h)Webhook worker
Dead-letter thresholdSubscription disabled after 20 consecutive failuresWebhook worker
Audit log retention, Private30 daysPlan matrix
Audit log retention, Pro90 daysPlan matrix
Audit log retention, Business and Business API24 months (730 days)Plan matrix
Audit log retention, Enterprise5 years (1825 days)Plan matrix
OAuth code validity10 minutesOAuth server
Magic link validity15 minutesAuth service
API token formatkmt_ + 43 base64url characters (256 bits of entropy)Token module
Maximum tokens per workspace50 active (more on request for Enterprise)Token module

Behaviour on exceedance

  • Rate limit (429): response carries a Retry-After header in seconds. Clients should wait and retry, ideally with jitter.
  • Memory limit (402): the call is rejected; the response indicates the required plan. Existing content stays untouched.
  • Document too large (413): the call is rejected. Split the content and ingest in multiple calls.
  • Webhook dead letter: the delivery is marked dead. After 20 consecutive failures the whole subscription is disabled and must be reactivated manually in the dashboard.

Related