name: Tailcall / ForgeCode Rate Limits
description: >-
  Rate limit information for Tailcall's products. The open-source GraphQL
  runtime has no platform-imposed rate limits — limits are solely determined
  by the operator's infrastructure and upstream API providers. ForgeCode, the
  commercial CLI product, enforces daily AI request quotas per subscription tier.
version: "0.1"
created: "2026-06-13"
modified: "2026-06-13"

products:
  - name: Tailcall GraphQL Runtime (Self-Hosted)
    description: >-
      No rate limits imposed by Tailcall Inc. The runtime itself supports
      configurable upstream timeouts, connection pool sizes, and request
      batching windows via .graphql DSL directives. Operators set their own
      limits based on infrastructure capacity and upstream API constraints.
    platformLimits: none
    configurableControls:
      - name: "@upstream timeout"
        description: Per-upstream HTTP request timeout in milliseconds
        directive: "@upstream(timeout: <ms>)"
      - name: "@upstream poolSize"
        description: HTTP connection pool size per upstream
        directive: "@upstream(poolSize: <n>)"
      - name: "@batch"
        description: Batch window and size for N+1 query coalescing
        directive: "@batch(delay: <ms>, maxSize: <n>)"
      - name: "@cache"
        description: Response cache TTL for upstream HTTP calls
        directive: "@cache(maxAge: <seconds>)"
    notes: >-
      Rate limits from upstream REST/gRPC/GraphQL APIs still apply. Tailcall
      passes through HTTP 429 responses from upstreams. Operators should
      configure appropriate retry and circuit-breaker logic at the infrastructure
      layer (e.g., AWS API Gateway throttling, Cloudflare rate limiting).

  - name: ForgeCode Free Tier
    description: Dynamic daily AI request limit based on server load.
    limits:
      - scope: daily
        quantity: "10–50"
        unit: AI requests
        enforcement: soft (dynamic, load-dependent)
        resetPeriod: 24 hours (rolling)
    notes: >-
      The free tier limit fluctuates with overall platform load. No hard SLA
      is provided for free users during peak demand periods.

  - name: ForgeCode Pro Tier
    description: Fixed daily AI request quota for Pro subscribers.
    limits:
      - scope: daily
        quantity: 1000
        unit: AI requests
        enforcement: hard
        resetPeriod: 24 hours (rolling)
    notes: Pro users receive priority access during high server load periods.

  - name: ForgeCode Max Tier
    description: Fixed daily AI request quota for Max subscribers.
    limits:
      - scope: daily
        quantity: 5000
        unit: AI requests
        enforcement: hard
        resetPeriod: 24 hours (rolling)
    notes: Max tier receives highest throughput allocation on the platform.