specification: FinOps Framework
specificationVersion: '1.0'
schema: https://www.finops.org/framework/
provider: Langbase
providerId: langbase
created: '2026-07-01'
modified: '2026-07-01'
reconciled: false
tags:
- AI
- Agents
- RAG
- LLM
- Serverless
- FinOps
- Cost Management
- FOCUS
description: >-
  FinOps view of Langbase spend. Langbase blends subscription tiers (Free, Pro,
  Enterprise) with usage-based metering across its primitives: Pipe / Agent
  runs, Memory storage and retrieval, Parser document parsing, Chunker, Embed
  (per embedding model), and Tools (web search / crawl). When customers bring
  their own LLM provider key via the LB-LLM-Key header, the underlying model
  token spend is billed by the downstream provider (OpenAI, Anthropic, Google,
  and others) rather than by Langbase, splitting cost across two invoices.
notes: >-
  Per-unit rates and included allotments are subject to change; verify against
  the Langbase pricing page during reconciliation.
sources:
- https://langbase.com/pricing
- https://langbase.com/docs
- https://focus.finops.org/focus-specification/v1-3/
alignedWith:
  framework: FinOps Foundation Framework
  frameworkUrl: https://www.finops.org/framework/
  dataSpec: FOCUS
  dataSpecVersion: '1.3'
  dataSpecUrl: https://focus.finops.org/focus-specification/v1-3/
publisherName: Langbase
serviceCategory: AI and Machine Learning
billingModel:
  pricingCategory: Usage-Based
  billingFrequency: Monthly
  billingCurrency: USD
  chargeCategories:
  - Usage
  - Purchase
  - Adjustment
focusColumns:
  ServiceName: Langbase
  ServiceCategory: AI and Machine Learning
  ProviderName: Langbase
  PublisherName: Langbase
  InvoiceIssuerName: Langbase
  BillingCurrency: USD
  ChargeCategory: Usage
  PricingCategory: Usage-Based
meters:
- name: agent_pipe_runs
  description: Executions of Pipes and Agent runs, metered per run.
  unit: runs
  aggregation: sum
  dimensions:
  - account
  - pipe
  - model
- name: memory_retrievals
  description: Memory retrieval (RAG) requests, metered per request.
  unit: requests
  aggregation: sum
  dimensions:
  - account
  - memory
- name: memory_storage
  description: Documents stored and embedded in memory stores.
  unit: documents
  aggregation: sum
  dimensions:
  - account
  - memory
- name: embedding_tokens
  description: Tokens embedded via the Embed API, per embedding model.
  unit: tokens
  aggregation: sum
  dimensions:
  - account
  - model
- name: parser_documents
  description: Documents parsed via the Parser API (up to 10 MB each).
  unit: documents
  aggregation: sum
  dimensions:
  - account
- name: tool_invocations
  description: Tool calls (web search, crawl) metered per invocation.
  unit: invocations
  aggregation: sum
  dimensions:
  - account
  - tool
- name: subscription
  description: Monthly Pro subscription fee, or Enterprise agreement.
  unit: months
  aggregation: sum
  dimensions:
  - account
  - tier
- name: byok_llm_tokens
  description: >-
    Underlying LLM tokens billed by the downstream provider when using
    bring-your-own-key (LB-LLM-Key). Not billed by Langbase.
  unit: tokens
  aggregation: sum
  dimensions:
  - account
  - provider
  - model
principles:
- name: Visibility
  description: Track Langbase run, retrieval, and tool usage per API key; separately track downstream LLM token spend when using bring-your-own-key.
- name: Allocation
  description: Scope API keys per Pipe / workload / team and map to internal cost centers.
- name: Optimization
  description: Right-size models via the provider:model_id string; tune topK and chunk sizes to cut embedding and retrieval cost; cache and reuse memory; batch embed up to 100 chunks per request.
- name: Accountability
  description: Assign owners per Pipe / project; review run, retrieval, and BYOK token spend monthly against budget.
maintainers:
- FN: Kin Lane
  email: kin@apievangelist.com