Budget Controls

AI agents consume LLM tokens on every request, and token consumption translates directly to cost. Without budget controls, a single agent caught in a reasoning loop or processing unexpectedly large inputs can exhaust thousands of dollars in minutes. VeraID provides multi-tier budget enforcement that checks spend limits before tokens are consumed, preventing runaway costs at the source.

Budget Types

Every AI agent identity supports four budget controls, each independently configurable.

Budget Type	Description	Example
Daily limit	Maximum spend per UTC day. Resets at midnight UTC.	$50.00/day
Monthly limit	Maximum spend per calendar month. Resets on the 1st of each month at midnight UTC.	$1,000.00/month
Per-request limit	Maximum spend for a single LLM call. Prevents individual requests from consuming disproportionate budget.	$2.00/request
Warning threshold	Percentage of any limit at which alerts are triggered. Does not block requests.	80%

Note

All budget amounts are denominated in USD. Token costs are calculated using the provider’s published pricing for the specific model at the time of the request.

---

Action on Exceed

When a budget limit is reached, VeraID takes one of three configurable actions.

Block

The request is rejected immediately with a 429 Budget Exceeded response. The agent receives a structured error indicating which limit was exceeded and when the budget resets. This is the default and recommended action.

{
  "error": "BUDGET_EXCEEDED",
  "message": "Daily budget limit exceeded",
  "details": {
    "limit": 50.00,
    "spent": 49.87,
    "estimatedCost": 0.45,
    "resetsAt": "2026-03-20T00:00:00Z"
  }
}

Throttle

Requests are queued rather than rejected. The gateway holds the request and retries it after the budget resets or after an administrator manually increases the limit. Throttled requests have a configurable timeout (default: 30 minutes) after which they are dropped.

Alert Only

The request is allowed to proceed, but an alert is sent to configured notification channels (email, Slack, webhook). Use this action during initial rollout to understand spending patterns before enforcing hard limits.

Caution

alert-only mode provides no cost protection. Use it only during an observation period to calibrate budget limits. Switch to block or throttle before moving agents to production.

---

Rate Limiting

In addition to cost-based budgets, VeraID enforces request rate limits to prevent agents from overwhelming LLM providers or consuming budget through high-frequency low-cost calls.

Rate Limit	Description	Default
Requests per minute	Maximum LLM API calls per 60-second sliding window	60
Requests per hour	Maximum LLM API calls per 60-minute sliding window	1,000

Rate limits are enforced independently of budget limits. A request can be within budget but still rate-limited, or within rate limits but over budget.

{
  "agentConfig": {
    "tokenBudget": {
      "daily": 50.00,
      "monthly": 1000.00,
      "perRequest": 2.00
    },
    "rateLimits": {
      "requestsPerMinute": 30,
      "requestsPerHour": 500
    }
  }
}

---

Creating an Agent with Budget Controls

API

curl -X POST https://app.veraid.io/api/v1/identities \
  -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "research-assistant",
    "type": "AI_AGENT",
    "description": "Summarizes research papers and answers questions from the knowledge base",
    "tags": ["research", "production"],
    "agentConfig": {
      "provider": "anthropic",
      "modelIdentifier": "claude-sonnet-4",
      "capabilities": ["chat-completion", "retrieval"],
      "dataAccess": {
        "allowed": ["knowledge-base", "public-papers"],
        "denied": ["financial-records", "hr-data"]
      },
      "monitoring": {
        "logAllRequests": true,
        "logResponses": false,
        "requireApproval": false
      },
      "tokenBudget": {
        "daily": 25.00,
        "monthly": 500.00,
        "perRequest": 1.50,
        "warningThreshold": 0.80,
        "onExceed": "block"
      },
      "rateLimits": {
        "requestsPerMinute": 20,
        "requestsPerHour": 300
      }
    }
  }'

SDK

import { VeraIDClient } from '@veraid/sdk';

const client = new VeraIDClient({
  apiKey: process.env.VERAID_API_KEY,
  agentId: process.env.VERAID_AGENT_ID,
});

const agent = await client.identities.create({
  name: 'research-assistant',
  type: 'AI_AGENT',
  description: 'Summarizes research papers and answers questions from the knowledge base',
  tags: ['research', 'production'],
  agentConfig: {
    provider: 'anthropic',
    modelIdentifier: 'claude-sonnet-4',
    capabilities: ['chat-completion', 'retrieval'],
    dataAccess: {
      allowed: ['knowledge-base', 'public-papers'],
      denied: ['financial-records', 'hr-data'],
    },
    monitoring: {
      logAllRequests: true,
      logResponses: false,
      requireApproval: false,
    },
    tokenBudget: {
      daily: 25.0,
      monthly: 500.0,
      perRequest: 1.5,
      warningThreshold: 0.8,
      onExceed: 'block',
    },
    rateLimits: {
      requestsPerMinute: 20,
      requestsPerHour: 300,
    },
  },
});

---

Checking Budget at Runtime

Check Before Spending

Use checkBudget before making an LLM call to verify that the estimated cost is within limits. This prevents wasted computation on requests that would be rejected at the gateway.

// Estimate the cost of the upcoming request
const estimatedCost = 0.03;

const budgetCheck = await client.agents.checkBudget(estimatedCost);

if (!budgetCheck.allowed) {
  console.error(`Budget check failed: ${budgetCheck.reason}`);
  console.error(`Resets at: ${budgetCheck.resetsAt}`);
  // Gracefully degrade or queue for later
  return;
}

// Proceed with the LLM call
const response = await client.agents.complete({
  model: 'claude-sonnet-4',
  messages: [{ role: 'user', content: userInput }],
});

The checkBudget response:

{
  "allowed": true,
  "estimatedCost": 0.03,
  "daily": {
    "limit": 25.00,
    "spent": 12.47,
    "remaining": 12.53
  },
  "monthly": {
    "limit": 500.00,
    "spent": 187.23,
    "remaining": 312.77
  }
}

Get Budget Status

Use getBudgetStatus to retrieve the current budget state without performing a check against a specific cost.

const status = await client.agents.getBudgetStatus();

console.log(`Daily: $${status.daily.spent} / $${status.daily.limit}`);
console.log(`Monthly: $${status.monthly.spent} / $${status.monthly.limit}`);
console.log(`Requests today: ${status.daily.requestCount}`);
console.log(`Warning triggered: ${status.warningTriggered}`);

{
  "daily": {
    "limit": 25.00,
    "spent": 12.47,
    "remaining": 12.53,
    "requestCount": 142,
    "resetsAt": "2026-03-20T00:00:00Z"
  },
  "monthly": {
    "limit": 500.00,
    "spent": 187.23,
    "remaining": 312.77,
    "requestCount": 4218,
    "resetsAt": "2026-04-01T00:00:00Z"
  },
  "warningThreshold": 0.80,
  "warningTriggered": false,
  "onExceed": "block"
}

---

Cost Tracking

The gateway automatically tracks cost data for every LLM call made through a governed agent identity.

Per-Request Tracking

Each request records:

Field	Description
inputTokens	Number of tokens in the prompt
outputTokens	Number of tokens in the response
totalTokens	Sum of input and output tokens
model	Model identifier used for the request
costUsd	Calculated cost in USD based on model pricing
timestamp	ISO 8601 timestamp of the request
operation	Operation type (e.g., chat-completion, embedding, image-generation)

Cumulative Tracking

VeraID maintains running totals at multiple granularities:

• Per-request — Individual cost for each LLM call
• Hourly — Aggregated spend per clock hour (for trend analysis)
• Daily — Aggregated spend per UTC day (for daily budget enforcement)
• Monthly — Aggregated spend per calendar month (for monthly budget enforcement)
• Lifetime — Total spend since the agent identity was created

Cost Analytics API

# Get cost breakdown for an agent over a time range
curl -G https://app.veraid.io/api/v1/identities/{agentId}/costs \
  -H "Authorization: Bearer $API_KEY" \
  -d "startDate=2026-03-01" \
  -d "endDate=2026-03-19" \
  -d "granularity=daily"

{
  "agentId": "idt_9c3a8b7e-4f21-4d6a-b8e1-a2c5d9f07e3b",
  "period": {
    "start": "2026-03-01T00:00:00Z",
    "end": "2026-03-19T23:59:59Z"
  },
  "totalCost": 187.23,
  "totalRequests": 4218,
  "totalTokens": 2847392,
  "byDay": [
    {
      "date": "2026-03-01",
      "cost": 8.42,
      "requests": 198,
      "inputTokens": 52340,
      "outputTokens": 31204
    }
  ],
  "byModel": [
    {
      "model": "claude-sonnet-4",
      "cost": 142.18,
      "requests": 3102,
      "percentage": 75.9
    },
    {
      "model": "gpt-4o",
      "cost": 45.05,
      "requests": 1116,
      "percentage": 24.1
    }
  ]
}

Tip

Use the cost analytics API to right-size your budget limits. Start with generous limits and alert-only mode for the first week, analyze actual spend patterns, then tighten limits to 120-150% of observed peak daily spend.

---

What’s Next

• Prompt Injection Detection — Protect agents from adversarial inputs
• Approval Workflows — Require human review for high-cost operations
• Agent Gateway — Understand the full request pipeline