Clone OpenAI pricing

In this article, you will learn how to build a billing system with Lago based on tokens. This template is suitable for Large Language Model (LLM) and Generative AI companies whose pricing can vary based on the application or model used.

Pricing structure

For OpenAI, pricing depends on the language model used. Here are several price points they offer: “Prices are per 1,000 tokens. You can think of tokens as pieces of words, where 1,000 tokens is about 750 words (learn more here).”

Model	Input	Output
8K context	$0.03 / 1,000 tokens	$0.06 / 1,000 tokens
32K context	$0.06 / 1,000 tokens	$0.12 / 1,000 tokens

Get started

Set up Lago

LAGO_URL="https://api.getlago.com"
API_KEY="__API_KEY__"

from lago_python_client.client import Client
from lago_python_client.exceptions import LagoApiError

client = Client(api_key='__API_KEY__')

require 'lago-ruby-client'

client = Lago::Api::Client.new(api_key: '__API_KEY__')

import { Client } from 'lago-javascript-client';

const client = Client('__API_KEY__');

package main

import (
    "context"
    lago "github.com/getlago/lago-go-client"
)

client := lago.New().SetApiKey("__API_KEY__")

ctx := context.TODO()

Create a billable metric

Create a single billable metric to track token usage across different models and input/output types.

Add a unique code for your billable metric
Set aggregation_type to sum_agg
Set field_name to tokens
Set recurring to false
Set filters to distinguish between models and token types

curl --location --request POST "$LAGO_URL/api/v1/billable_metrics" \
  --header "Authorization: Bearer $API_KEY" \
  --header 'Content-Type: application/json' \
  --data-raw '{
    "billable_metric": {
      "name": "AI Tokens",
      "code": "__BILLABLE_METRIC_CODE__",
      "description": "Token usage across OpenAI models",
      "aggregation_type": "sum_agg",
      "field_name": "tokens",
      "recurring": false,
      "filters": [
        {
          "key": "model",
          "values": ["8k", "32k"]
        },
        {
          "key": "type", 
          "values": ["input", "output"]
        }
      ]
    }
  }'

from lago_python_client.models import BillableMetric, BillableMetricFilter

model_filter = BillableMetricFilter(
  key='model',
  values=['8k', '32k']
)

type_filter = BillableMetricFilter(
  key='type',
  values=['input', 'output']
)

billable_metric = BillableMetric(
  name='Tokens',
  code='__BILLABLE_METRIC_CODE__',
  description='Token usage across OpenAI models',
  aggregation_type='sum_agg',
  field_name='tokens',
  recurring=False,
  filters=[model_filter, type_filter]
)

client.billable_metrics.create(billable_metric)

client.billable_metrics.create({
  name: 'Tokens',
  code: '__BILLABLE_METRIC_CODE__', 
  description: 'Token usage across OpenAI models',
  aggregation_type: 'sum_agg',
  field_name: 'tokens',
  recurring: false,
  filters: [
    {
      key: 'model',
      values: ['8k', '32k']
    },
    {
      key: 'type',
      values: ['input', 'output']
    }
  ]
})

const billableMetric = {
  name: "Tokens",
  code: "__BILLABLE_METRIC_CODE__",
  description: "Token usage across OpenAI models", 
  aggregation_type: "sum_agg",
  field_name: "tokens",
  recurring: false,
  filters: [
    {
      key: "model",
      values: ["8k", "32k"]
    },
    {
      key: "type",
      values: ["input", "output"]
    }
  ]
};

await client.billableMetrics.createBillableMetric(billableMetric);

bmInput := &lago.BillableMetricInput{
  Name:            "Tokens",
  Code:            "__BILLABLE_METRIC_CODE__",
  Description:     "Token usage across OpenAI models",
  AggregationType: lago.SumAggregation,
  FieldName:       "tokens",
  Recurring:       false,
  Filters:         []lago.BillableMetricFilter{
    {
      Key:    "type",
      Values: []string{"input", "output"},
    },
    {
      Key:    "model",
      Values: []string{"8k", "32k"},
    },
  },
}

billableMetric, err := lagoClient.BillableMetric().Create(bmInput)

Refer to the API reference and guide on filters to learn more.

Create a plan

Create a plan to price packages of tokens used with OpenAI’s specific pricing tiers.

Add a unique code for your plan
Under charges, configure pricing tiers for different models and token types

curl --location --request POST "$LAGO_URL/api/v1/plans" \
  --header "Authorization: Bearer $API_KEY" \
  --header 'Content-Type: application/json' \
  --data-raw '{
    "plan": {
      "name": "OpenAI Pricing",
      "code": "__PLAN_CODE__",
      "amount_cents": 0,
      "amount_currency": "USD",
      "interval": "monthly",
      "pay_in_advance": false,
      "charges": [
        {
          "billable_metric_id": "__BILLABLE_METRIC_ID__",
          "charge_model": "package",
          "filters": [
            {
              "invoice_display_name": "8k-input",
              "properties": {
                "amount": "0.03",
                "free_units": 0,
                "package_size": 1000000
              },
              "values": {
                "model": ["8k"],
                "type": ["input"]
              }
            },
            {
              "invoice_display_name": "8k-output", 
              "properties": {
                "amount": "0.06",
                "free_units": 0,
                "package_size": 1000000
              },
              "values": {
                "model": ["8k"],
                "type": ["output"]
              }
            },
            {
              "invoice_display_name": "32k-input",
              "properties": {
                "amount": "0.06", 
                "free_units": 0,
                "package_size": 1000000
              },
              "values": {
                "model": ["32k"],
                "type": ["input"]
              }
            },
            {
              "invoice_display_name": "32k-output",
              "properties": {
                "amount": "0.12",
                "free_units": 0, 
                "package_size": 1000000
              },
              "values": {
                "model": ["32k"],
                "type": ["output"]
              }
            } 
          ]
        }
      ]
    }
  }'

from lago_python_client.models import Plan, Charge, ChargeFilter

# Filters for different models and token types
small_input_filter = ChargeFilter(
  invoice_display_name="8k-input",
  properties={"amount": "0.03", "free_units": 0, "package_size": 1000000},
  values={"model": ["8k"], "type": ["input"]}
)

small_output_filter = ChargeFilter(
  invoice_display_name="8k-output", 
  properties={"amount": "0.06", "free_units": 0, "package_size": 1000000},
  values={"model": ["8k"], "type": ["output"]}
)

large_input_filter = ChargeFilter(
  invoice_display_name="32k-input",
  properties={"amount": "0.06", "free_units": 0, "package_size": 1000000},
  values={"model": ["32k"], "type": ["input"]}
)

large_output_filter = ChargeFilter(
  invoice_display_name="32k-output",
  properties={"amount": "0.12", "free_units": 0, "package_size": 1000000},
  values={"model": ["32k"], "type": ["output"]}
)

# Charge for all filters
charge = Charge(
  billable_metric_id="__BILLABLE_METRIC_ID__",
  charge_model="package",
  filters=[small_input_filter, small_output_filter, large_input_filter, large_output_filter]
)

plan = Plan(
  name="OpenAI Plan",
  code="__PLAN_CODE__",
  amount_cents=0,
  amount_currency="USD", 
  interval="monthly",
  pay_in_advance=False,
  charges=[charge]
)

client.plans.create(plan)

client.plans.create({
  name: "OpenAI Pricing",
  code: "__PLAN_CODE__",
  amount_cents: 0,
  amount_currency: "USD",
  interval: "monthly",
  pay_in_advance: false,
  charges: [
    {
      billable_metric_id: "__BILLABLE_METRIC_ID__",
      charge_model: "package",
      filters: [
        {
          invoice_display_name: "8k-input",
          properties: { amount: "0.03", free_units: 0, package_size: 1000000 },
          values: { model: ["8k"], type: ["input"] }
        },
        {
          invoice_display_name: "8k-output",
          properties: { amount: "0.06", free_units: 0, package_size: 1000000 },
          values: { model: ["8k"], type: ["output"] }
        },
        {
          invoice_display_name: "32k-input", 
          properties: { amount: "0.06", free_units: 0, package_size: 1000000 },
          values: { model: ["32k"], type: ["input"] }
        },
        {
          invoice_display_name: "32k-output",
          properties: { amount: "0.12", free_units: 0, package_size: 1000000 },
          values: { model: ["32k"], type: ["output"] }
        }
      ]
    }
  ]
})

const plan = {
  name: "OpenAI Pricing",
  code: "__PLAN_CODE__",
  amount_cents: 0,
  amount_currency: "USD",
  interval: "monthly",
  pay_in_advance: false,
  charges: [
    {
      billable_metric_id: "__BILLABLE_METRIC_ID__",
      charge_model: "package",
      filters: [
        {
          invoice_display_name: "8k-input",
          properties: { amount: "0.03", free_units: 0, package_size: 1000000 },
          values: { model: ["8k"], type: ["input"] }
        },
        {
          invoice_display_name: "8k-output",
          properties: { amount: "0.06", free_units: 0, package_size: 1000000 },
          values: { model: ["8k"], type: ["output"] }
        },
        {
          invoice_display_name: "32k-input",
          properties: { amount: "0.06", free_units: 0, package_size: 1000000 },
          values: { model: ["32k"], type: ["input"] }
        },
        {
          invoice_display_name: "32k-output", 
          properties: { amount: "0.12", free_units: 0, package_size: 1000000 },
          values: { model: ["32k"], type: ["output"] }
        }
      ]
    }
  ]
};

await client.plans.createPlan({ plan });

planInput := &lago.PlanInput{
  Name:           "OpenAI Pricing",
  Code:           "__PLAN_CODE__", 
  AmountCents:    0,
  AmountCurrency: "USD",
  Interval:       "monthly",
  PayInAdvance:   false,
  Charges:        []lago.PlanChargeInput{
    {
      BillableMetricID: "__BILLABLE_METRIC_ID__",
      ChargeModel:        "package",
      Filters: []lago.ChargeFilter{
        {
          InvoiceDisplayName: "8k-input",
          Properties: map[string]interface{}{
            "amount": "0.03", "free_units": 0, "package_size": 1000000,
          },
          Values: map[string]string{
            "model": []string{"8k"}, "type": []string{"input"},
          },
        },
        {
          InvoiceDisplayName: "8k-output",
          Properties: map[string]interface{}{
            "amount": "0.06", "free_units": 0, "package_size": 1000000,
          },
          Values: map[string]string{
            "model": []string{"8k"}, "type": []string{"output"},
          },
        },
        {
          InvoiceDisplayName: "32k-input",
          Properties: map[string]interface{}{
            "amount": "0.06", "free_units": 0, "package_size": 1000000,
          },
          Values: map[string]string{
            "model": []string{"32k"}, "type": []string{"input"},
          },
        },
        {
          InvoiceDisplayName: "32k-output",
          Properties: map[string]interface{}{
            "amount": "0.12", "free_units": 0, "package_size": 1000000,
          },
          Values: map[string]string{
            "model": []string{"32k"}, "type": []string{"output"},
          },
        },
      },
    },
  },
}

plan, err := lagoClient.Plan().Create(planInput)

Refer to the API reference and guide on charges with filters to learn more.

Create a customer

Create a customer with a unique external_id.

curl --location --request POST "$LAGO_URL/api/v1/customers" \
  --header "Authorization: Bearer $API_KEY" \
  --header 'Content-Type: application/json' \
  --data-raw '{
    "customer": {
      "external_id": "__EXTERNAL_CUSTOMER_ID__",
      "name": "Acme Inc",
      "email": "john@acme.com",
      "currency": "USD",
      "timezone": "America/New_York"
    }
  }'

from lago_python_client.models import Customer

customer = Customer(
  external_id="__EXTERNAL_CUSTOMER_ID__",
  name="Acme Inc",
  email="john@acme.com",
  currency="USD",
  timezone="America/New_York"
)

client.customers.create(customer)

client.customers.create(
  external_id: "__EXTERNAL_CUSTOMER_ID__",
  name: "Acme Inc", 
  email: "john@acme.com",
  currency: "USD",
  timezone: "America/New_York"
)

const customer = {
  external_id: "__EXTERNAL_CUSTOMER_ID__",
  name: "Acme Inc",
  email: "john@acme.com",
  currency: "USD",
  timezone: "America/New_York"
};

await client.customers.createCustomer({ customer });

customerInput := &lago.CustomerInput{
  ExternalID: "__EXTERNAL_CUSTOMER_ID__",
  Name:       "Acme Inc",
  Email:      "john@acme.com",
  Currency:   "USD",
  Timezone:   "America/New_York",
}

customer, err := client.Customer().Create(ctx, customerInput)

Refer to the API reference to create a customer.

Create a subscription

Create a subscription for the customer with the plan’s code.

curl --location --request POST "$LAGO_URL/api/v1/subscriptions" \
  --header "Authorization: Bearer $API_KEY" \
  --header 'Content-Type: application/json' \
  --data-raw '{
    "subscription": {
      "external_customer_id": "__EXTERNAL_CUSTOMER_ID__",
      "plan_code": "__PLAN_CODE__",
      "external_id": "__EXTERNAL_SUBSCRIPTION_ID__"
    }
  }'

from lago_python_client.models import Subscription

subscription = Subscription(
  external_customer_id="__EXTERNAL_CUSTOMER_ID__",
  plan_code="__PLAN_CODE__",
  external_id="__EXTERNAL_SUBSCRIPTION_ID__"
)

client.subscriptions.create(subscription)

client.subscriptions.create(
  external_customer_id: "__EXTERNAL_CUSTOMER_ID__",
  plan_code: "__PLAN_CODE__",
  external_id: "__EXTERNAL_SUBSCRIPTION_ID__"
)

const subscription = {
  external_customer_id: "__EXTERNAL_CUSTOMER_ID__",
  plan_code: "__PLAN_CODE__",
  external_id: "__EXTERNAL_SUBSCRIPTION_ID__"
};

await client.subscriptions.createSubscription({ subscription });

subscriptionInput := &lago.SubscriptionInput{
  ExternalCustomerID: "__EXTERNAL_CUSTOMER_ID__",
  PlanCode:           "__PLAN_CODE__",
  ExternalID:         "__EXTERNAL_SUBSCRIPTION_ID__",
}

subscription, err := client.Subscription().Create(ctx, subscriptionInput)

Refer to the API reference to create a subscription.

Ingest usage via events

Send usage events to Lago to track usage.

Reference your billable metric with code
Reference the customer’s subscription with external_subscription_id
Include usage and filters data in properties

LAGO_URL="https://api.getlago.com"
API_KEY="__API_KEY__"

curl --location --request POST "$LAGO_URL/api/v1/events" \
  --header "Authorization: Bearer $API_KEY" \
  --header 'Content-Type: application/json' \
  --data-raw '{
    "event": {
      "transaction_id": "__TRANSACTION_ID__",
      "code": "__BILLABLE_METRIC_CODE__",
      "external_subscription_id": "__EXTERNAL_SUBSCRIPTION_ID__",
      "properties": {
        "total": 5000,
        "model": "8k",
        "type": "input"
      }
    }
  }'

from lago_python_client.client import Client
from lago_python_client.exceptions import LagoApiError
from lago_python_client.models import Event

client = Client(api_key='__API_KEY__')

event = Event(
  transaction_id="__TRANSACTION_ID__",
  code="__BILLABLE_METRIC_CODE__",
  external_subscription_id="__EXTERNAL_SUBSCRIPTION_ID__",
  properties={
    "total": 5000,
    "model": "8k",
    "type": "input"
  }
)

try:
    client.events.create(event)
except LagoApiError as e:
    repair_broken_state(e)  # do something on error or raise your own exception

require 'lago-ruby-client'

client = Lago::Api::Client.new(api_key: '__API_KEY__')

client.events.create(
  transaction_id: "__TRANSACTION_ID__",
  code: "__BILLABLE_METRIC_CODE__",
  external_subscription_id: "__EXTERNAL_SUBSCRIPTION_ID__",
  properties: {
    total: 5000,
    model: "8k",
    type: "input"
  }
)

await client.events.createEvent({
  event: {
    transaction_id: "__TRANSACTION_ID__",
    code: "__BILLABLE_METRIC_CODE__",
    external_subscription_id: "__EXTERNAL_SUBSCRIPTION_ID__",
    properties: {
      total: 5000,
      model: "8k",
      type: "input"
    }
  }
});

eventInput := &lago.EventInput{
  TransactionID:          "__TRANSACTION_ID__",
  Code:                   "__BILLABLE_METRIC_CODE__",
  ExternalSubscriptionID: "__EXTERNAL_SUBSCRIPTION_ID__",
  Properties: map[string]interface{}{
    "total": 5000,
    "model": "8k",
    "type":  "input",
  },
}

event, err := client.Event().Create(ctx, eventInput)

Refer to the API reference to create an event.

Monitor current usage

Track real-time customer usage for the current billing period.

curl --location --request GET "$LAGO_URL/api/v1/customers/__EXTERNAL_CUSTOMER_ID__/current_usage?external_subscription_id=__EXTERNAL_SUBSCRIPTION_ID__" \
--header 'Authorization: Bearer __API_KEY__' \
--header 'Content-Type: application/json'

customer_usage = client.customers.current_usage('__EXTERNAL_CUSTOMER_ID__', '__EXTERNAL_SUBSCRIPTION_ID__')

customer_usage = client.customer.current_usage('__EXTERNAL_CUSTOMER_ID__', '__EXTERNAL_SUBSCRIPTION_ID__')

const customerUsage = await client.customers.findCustomerCurrentUsage(
    '__EXTERNAL_CUSTOMER_ID__',
    { external_subscription_id: '__EXTERNAL_SUBSCRIPTION_ID__' }
)

customerUsage, err := client.Customer().CurrentUsage(ctx, "__EXTERNAL_CUSTOMER_ID__")

Refer to the API reference to get the current usage.

Wrap-up

Per-token pricing offers flexibility and visibility, and allows LLM and Generative AI companies like OpenAI to attract more customers. With Lago, you can create your own metric dimensions to adapt this template to your products and services. Give it a try, click here to get started!

​Pricing structure

​Get started

​Wrap-up

Pricing structure

Get started

Wrap-up