Rivano calculates the cost of every proxied request using your provider’s pricing model. You get per-agent cost breakdowns, budget alerts, and SLA monitoring without any code changes in your application.

Step 1: Configure provider pricing

Rivano uses standard pricing for major providers out of the box. If you use a custom deployment or negotiated pricing, update the provider record:

import Rivano from '@rivano/sdk';

const rivano = new Rivano({ apiKey: process.env.RIVANO_API_KEY! });

await rivano.providers.create({
  name: 'openai-custom',
  provider: 'openai',
  apiKey: process.env.OPENAI_API_KEY!,
  pricing: {
    'gpt-4o': {
      inputCostPer1kTokens: 0.0025,
      outputCostPer1kTokens: 0.01,
    },
    'gpt-4o-mini': {
      inputCostPer1kTokens: 0.00015,
      outputCostPer1kTokens: 0.0006,
    },
  },
});

curl -X POST https://api.rivano.ai/api/providers \
  -H "Authorization: Bearer rv_api_..." \
  -H "Content-Type: application/json" \
  -d '{
    "name": "openai-custom",
    "provider": "openai",
    "apiKey": "sk-...",
    "pricing": {
      "gpt-4o": {
        "inputCostPer1kTokens": 0.0025,
        "outputCostPer1kTokens": 0.01
      }
    }
  }'

Step 2: View cost breakdown

Cost data is available in Observability → Costs → Breakdown. The breakdown shows total spend grouped by agent, model, and time period.

Query cost data via the API:

const breakdown = await rivano.costs.breakdown({
  period: '30d',
  groupBy: 'agent',
});

for (const entry of breakdown.data) {
  console.log(`${entry.agentName}: $${entry.totalCostUsd.toFixed(4)}`);
}

curl "https://api.rivano.ai/api/costs/breakdown?period=30d&groupBy=agent" \
  -H "Authorization: Bearer rv_api_..."

Step 3: Create budgets

Set a budget to get alerted when an agent approaches its spending limit:

await rivano.costs.createBudget({
  name: 'customer-support-monthly',
  agentId: 'agent_abc123',
  period: 'monthly',
  limitUsd: 500,
});

curl -X POST https://api.rivano.ai/api/costs/budgets \
  -H "Authorization: Bearer rv_api_..." \
  -H "Content-Type: application/json" \
  -d '{
    "name": "customer-support-monthly",
    "agentId": "agent_abc123",
    "period": "monthly",
    "limitUsd": 500
  }'

Rivano sends an alert when usage reaches 80% and 100% of the budget. At 100%, you can optionally configure the budget to block further requests from that agent for the remainder of the period.

💡

Create a budget for each production agent. It takes 30 seconds and prevents surprise bills.

Step 4: Set up alerts

Configure where budget and SLA alerts are delivered:

# Via the dashboard: Settings → Alerts → + Alert Channel
# Choose: email, webhook, or Slack webhook

# Via API (webhook example):
curl -X POST https://api.rivano.ai/api/alert-channels \
  -H "Authorization: Bearer rv_api_..." \
  -H "Content-Type: application/json" \
  -d '{
    "name": "ops-webhook",
    "type": "webhook",
    "url": "https://hooks.example.com/rivano-alerts",
    "events": ["budget.threshold_80", "budget.threshold_100", "sla.violation"]
  }'

Step 5: SLA monitoring

Create SLA targets for latency and error rate:

await rivano.costs.createSla({
  name: 'customer-support-sla',
  agentId: 'agent_abc123',
  p99LatencyMs: 3000,
  errorRatePercent: 1.0,
  evaluationPeriod: '1h',
});

curl -X POST https://api.rivano.ai/api/costs/sla \
  -H "Authorization: Bearer rv_api_..." \
  -H "Content-Type: application/json" \
  -d '{
    "name": "customer-support-sla",
    "agentId": "agent_abc123",
    "p99LatencyMs": 3000,
    "errorRatePercent": 1.0,
    "evaluationPeriod": "1h"
  }'

View active violations:

curl "https://api.rivano.ai/api/costs/sla/violations?period=24h" \
  -H "Authorization: Bearer rv_api_..."

Worked example: monitor a new agent

import Rivano from '@rivano/sdk';

const rivano = new Rivano({ apiKey: process.env.RIVANO_API_KEY! });

// 1. Register the agent
const agent = await rivano.agents.create({
  name: 'invoice-processor',
  modelProvider: 'anthropic',
  modelName: 'claude-opus-4-5',
  environment: 'production',
});

// 2. Set a monthly budget
await rivano.costs.createBudget({
  name: 'invoice-processor-budget',
  agentId: agent.data.id,
  period: 'monthly',
  limitUsd: 200,
});

// 3. Set an SLA target
await rivano.costs.createSla({
  name: 'invoice-processor-sla',
  agentId: agent.data.id,
  p99LatencyMs: 5000,
  errorRatePercent: 2.0,
  evaluationPeriod: '1h',
});

console.log('Agent monitoring configured');

Costs API — Full REST reference for budgets and SLA
Observability Dashboard — Costs page walkthrough
SDK Costs — Full SDK costs resource reference

Cost Tracking