specificationVersion: "0.1"
id: recall-ai-rate-limits
name: Recall.ai Rate Limits
description: >
  Recall.ai does not impose explicit per-request rate limits on its REST API as documented publicly. There are no additional charges for individual API requests. Usage limits are governed by active bot and recording time, which is billed per second. Bot variants (standard, web_4_core, web_gpu) have different infrastructure tiers. Developers should use webhooks rather than polling endpoints for bot status changes to avoid unnecessary API calls.
url: https://docs.recall.ai/docs/calculating-usage
limits:
  - scope: API Requests
    description: No documented hard rate limits on individual REST API calls. Recall.ai explicitly states there are no additional charges for individual API requests.
    limit: null
    unit: requests
    period: null
    notes: Use webhook events instead of polling for bot status changes in production.
  - scope: Meeting Bot - Standard
    description: Base meeting bot recording usage billed per second of active bot time from joining_call event to terminal event.
    limit: null
    unit: recording-hours
    period: per-second-billing
    costPerUnit: 0.50
    costCurrency: USD
    notes: Active time includes waiting_room and in_call_no_recording states.
  - scope: Meeting Bot - Web 4-Core Variant
    description: Premium bot variant with 4-core web infrastructure, adding $0.10/hr to the base recording rate.
    limit: null
    unit: recording-hours
    period: per-second-billing
    costPerUnit: 0.60
    costCurrency: USD
    notes: Base $0.50/hr + $0.10/hr premium = $0.60/hr total.
  - scope: Meeting Bot - Web GPU Variant
    description: Premium bot variant with GPU web infrastructure for high-compute workloads, adding $1.00/hr to the base recording rate.
    limit: null
    unit: recording-hours
    period: per-second-billing
    costPerUnit: 1.50
    costCurrency: USD
    notes: Base $0.50/hr + $1.00/hr premium = $1.50/hr total.
  - scope: Transcription - Recall.ai Built-in
    description: Built-in transcription billed per second based on difference between started_at and completed_at timestamps.
    limit: null
    unit: recording-hours
    period: per-second-billing
    costPerUnit: 0.15
    costCurrency: USD
  - scope: Storage
    description: Recording media storage with 7-day free tier; extended storage billed per media hour per 30-day period.
    limit: 7
    unit: days-free
    period: per-recording
    costPerUnit: 0.05
    costCurrency: USD
    notes: $0.05 per media hour for each additional 30-day period after the free 7 days.
  - scope: Regions
    description: Recall.ai operates four independent regional deployments. API keys and resources are region-scoped.
    regions:
      - name: US East
        baseURL: https://us-east-1.recall.ai
        alias: https://api.recall.ai
      - name: US West
        baseURL: https://us-west-2.recall.ai
      - name: EU Frankfurt
        baseURL: https://eu-central-1.recall.ai
      - name: Asia Tokyo
        baseURL: https://ap-northeast-1.recall.ai