specification: FinOps Framework
specificationVersion: '1.0'
provider: Agno
providerId: agno
url: https://www.agno.com/pricing
created: '2026-06-12'
modified: '2026-06-12'
description: >-
  Agno's billing model combines a flat-rate SaaS subscription for the AgentOS
  Control Plane (Pro at $150/month) with seat-based and connection-based
  add-ons. Underlying LLM API costs are pass-through from the customer's chosen
  model providers (OpenAI, Anthropic, Google, etc.), making total AI spend a
  combination of Agno platform fees plus LLM provider costs. The BYOC
  (Bring Your Own Cloud) deployment model means infrastructure costs are borne
  by the customer directly on their AWS, GCP, or Railway account.
billingModel:
  type: hybrid
  description: Flat monthly subscription plus seat/connection add-ons; LLM and infrastructure costs are external pass-through
  components:
    - name: Platform Subscription
      type: flat-rate
      description: Monthly fee for AgentOS Control Plane access (Pro tier)
      amount: 150.00
      currency: USD
      cadence: monthly
    - name: Seat Licenses
      type: per-unit
      description: Additional team member seats beyond the included 4
      amount: 30.00
      currency: USD
      unit: seat
      cadence: monthly
    - name: Live Connections
      type: per-unit
      description: Additional live AgentOS production connections beyond the included 1
      amount: 95.00
      currency: USD
      unit: connection
      cadence: monthly
    - name: LLM Provider Costs
      type: pass-through
      description: Token-based costs billed directly by LLM providers (OpenAI, Anthropic, Google, etc.) — not by Agno
      notes: Varies by provider, model, and token usage
    - name: Infrastructure (BYOC)
      type: pass-through
      description: Cloud infrastructure costs billed directly by AWS, GCP, or Railway — not by Agno
      notes: Customer manages their own cloud spend
focusColumns:
  - ChargeCategory
  - ChargeSubcategory
  - ServiceName
  - ServiceCategory
  - ResourceType
  - PricingUnit
  - BilledCost
  - EffectiveCost
  - BillingPeriodStart
  - BillingPeriodEnd
meters:
  - name: Platform Seat Hours
    description: Monthly active seats on the AgentOS Control Plane
    unit: seat/month
    dimension: seats
    granularity: monthly
  - name: Live Connections
    description: Active live connections to production AgentOS deployments
    unit: connection/month
    dimension: connections
    granularity: monthly
  - name: LLM Tokens (External)
    description: Tokens consumed from upstream LLM providers; tracked by Agno metrics API but billed externally
    unit: tokens
    dimension: llm_provider
    granularity: daily
    notes: Agno provides daily aggregated token analytics via the /metrics endpoint
  - name: Agent Runs
    description: Total agent and team runs executed through AgentOS
    unit: runs
    dimension: agent_id
    granularity: daily
principles:
  - name: Visibility
    description: >-
      Agno provides daily aggregated usage and token analytics through the
      AgentOS /metrics API endpoint, enabling teams to track LLM token
      consumption, run counts, and session activity. The Control Plane web
      console surfaces session monitoring and system metrics for operational
      visibility.
    actions:
      - Monitor token usage via the AgentOS metrics API (/metrics endpoint)
      - Use session monitoring in the Control Plane for run-level visibility
      - Track LLM provider costs separately through each provider's dashboard
      - Use trace inspection to identify expensive or inefficient agent patterns
  - name: Optimization
    description: >-
      Cost optimization in Agno involves right-sizing LLM model selection,
      using fallback models to avoid expensive retry loops, and leveraging
      caching for repeated LLM calls. The BYOC model allows customers to
      optimize infrastructure costs independently.
    actions:
      - Use fallback models to automatically switch to lower-cost models on rate limits
      - Implement LLM response caching to reduce redundant token usage
      - Limit chat history context to control per-run token costs
      - Use session summaries instead of full history to reduce context window usage
      - Right-size the AgentOS deployment infrastructure (BYOC) for actual workload
      - Monitor daily token analytics to identify high-cost agents or workflows