Docs/Core Concepts/Credits

Credit System

AgenticAnts uses a flexible, usage-based credit system that scales with your needs.

What are Credits?

Credits are AgenticAnts' unit of measurement for platform usage. Think of credits like cloud compute units - they represent the resources consumed by your AI operations.

Why Credits?

Flexibility: Use credits across all platform features
Predictability: Know costs before you commit
Scalability: Pay for what you use
Simplicity: One unit for all operations

How Credits Work

Credit Allocation

Credits are included with your plan and can be purchased additionally:

code

Pro Plan: $49/month ├─ 100,000 credits included ├─ Additional credits: $5 per 1,000 └─ No expiration within billing cycle Enterprise Plan: Custom pricing ├─ Custom credit allocation ├─ Volume discounts available └─ Dedicated support included

Credit Consumption

Different operations consume different amounts of credits:

Operation	Credits per Unit	Example
Trace Ingestion	1 per 1,000 traces	5,000 traces = 5 credits
Span Ingestion	0.1 per 1,000 spans	10,000 spans = 1 credit
Metric Data Points	0.05 per 1,000 points	20,000 points = 1 credit
Data Storage	5 per GB/month	10 GB = 50 credits/month
API Requests	0.5 per 1,000 requests	2,000 requests = 1 credit
Evaluation Runs	2 per 1,000 evaluations	500 evaluations = 1 credit

ℹ

Actual Usage: Your credit consumption depends on your actual usage patterns. Most teams use 5,000-15,000 credits per month.

Usage Examples

Example 1: Small Team

Profile: 5 developers, 3 AI agents, development and staging

code

Monthly Usage: ├─ Traces: 50,000 (50 credits) ├─ Spans: 200,000 (20 credits) ├─ Metrics: 100,000 (5 credits) ├─ Storage: 2 GB (10 credits) └─ API calls: 10,000 (5 credits) Total: 90 credits/month Cost: $49/month (Pro plan with 100,000 credits)

Example 2: Growing Startup

Profile: 20 developers, 10 AI agents, production workload

code

Monthly Usage: ├─ Traces: 500,000 (500 credits) ├─ Spans: 2,000,000 (200 credits) ├─ Metrics: 1,000,000 (50 credits) ├─ Storage: 20 GB (100 credits) └─ API calls: 100,000 (50 credits) Total: 900 credits/month Cost: $49 (Pro) + $0 (under 100,000 credits)

Example 3: Enterprise

Profile: 100+ developers, 50+ agents, high-volume production

code

Monthly Usage: ├─ Traces: 10,000,000 (10,000 credits) ├─ Spans: 50,000,000 (5,000 credits) ├─ Metrics: 20,000,000 (1,000 credits) ├─ Storage: 500 GB (2,500 credits) └─ API calls: 5,000,000 (2,500 credits) Total: 21,000 credits/month Cost: Custom Enterprise pricing

Monitoring Credit Usage

Real-Time Dashboard

View your credit consumption in real-time from the dashboard at https://app.agenticants.ai. The Usage view breaks down credits used, credits remaining, the reset date, and consumption by category (traces, metrics, storage). There is no programmatic usage API in the SDK; usage reporting is a dashboard feature.

Usage Alerts

Set up alerts for credit consumption in the dashboard at https://app.agenticants.ai. Under Usage settings you can configure threshold alerts (for example, warn at 70% and escalate at 90% of credits consumed) and route notifications to channels such as email, Slack, or PagerDuty. Alerts are a dashboard feature; the SDK does not expose an alerts API.

Usage Analytics

Analyze credit consumption patterns in the dashboard at https://app.agenticants.ai. The Usage analytics views show consumption trends over time, top consumers by project, and projected future usage. Metrics and trend queries are dashboard features and are not available through the SDK.

Credit Optimization

Tips to Reduce Credit Usage

1. Sampling

Don't trace everything - use intelligent sampling. Tracing is built on OpenTelemetry, so you sample by passing a standard OTel sampler to the NodeTracerProvider you register at startup (in an instrumentation.ts imported first). For example, head-based sampling of 10% of traces:

typescript

// instrumentation.ts const provider = new NodeTracerProvider({ // Sample 10% of traces (decided at trace start) sampler: new TraceIdRatioBasedSampler(0.1), spanProcessors: [ new AntsPlatformSpanProcessor({ publicKey: process.env.ANTS_PLATFORM_PUBLIC_KEY, secretKey: process.env.ANTS_PLATFORM_SECRET_KEY, baseUrl: "https://api.agenticants.ai", }), ], }); provider.register(); setAntsPlatformTracerProvider(provider);

2. Data Retention

Configure appropriate retention periods from the dashboard at https://app.agenticants.ai. Project settings let you set how long traces, metrics, logs, and raw data are kept. Retention is a dashboard/project setting, not an SDK API.

3. Batch Processing

The SDK batches span exports for you. The AntsPlatformSpanProcessor buffers observations and flushes them in batches, so instrumenting many operations does not produce one network call per operation:

typescript

// Each iteration records a span locally; the span processor // batches and exports them in the background. for (const item of items) { const span = startObservation("span", { name: "process-item" }); // ... do work ... span.end(); }

For short-lived scripts, flush the buffered batch before the process exits:

typescript

await provider.forceFlush(); // or: await provider.shutdown();

4. Compression

The SDK compresses and batches exports automatically over the wire. You only need to configure your credentials and host:

python

from ants_platform import AntsPlatform ants = AntsPlatform( public_key=os.getenv("ANTS_PLATFORM_PUBLIC_KEY"), secret_key=os.getenv("ANTS_PLATFORM_SECRET_KEY"), host="https://api.agenticants.ai", )

5. Selective Instrumentation

Only instrument what matters:

typescript

async function handle(request) { if (!shouldTrace(request)) { return process(request); // run without recording an observation } return startActiveObservation("span", { name: "handle-request" }, async (span) => { const result = await process(request); span.end(); return result; }); } function shouldTrace(request) { // Skip health checks if (request.path === "/health") return false; // Skip internal requests if (request.internal) return false; // Trace production traffic if (request.env === "production") return true; // Sample development traffic return Math.random() < 0.1; }

Purchasing Credits

Additional Credits

Need more credits? Purchase them as needed:

bash

# Via Dashboard 1. Go to Settings → Billing 2. Click "Purchase Credits" 3. Select quantity (1,000 to 1,000,000 credits) 4. Complete payment

Volume Discounts

Bulk purchases receive discounts:

Credits	Price per 1,000	Discount
1,000 - 9,999	$5.00	0%
10,000 - 49,999	$4.50	10%
50,000 - 99,999	$4.00	20%
100,000+	$3.50	30%

Enterprise Agreements

Custom credit packages for enterprise needs:

code

Enterprise Benefits: ├─ Custom credit allocation ├─ Flexible payment terms ├─ Volume discounts ├─ Dedicated support ├─ SLA guarantees └─ Custom retention policies

Credit Rollover

Monthly Plans

Credits reset at the start of each billing cycle
Unused credits do not roll over (Scale plan)
Purchase additional credits anytime

Enterprise Plans

Flexible rollover policies
Annual credit pools
Custom terms available

Fair Use Policy

AgenticAnts credits are subject to fair use:

Normal Use

Monitoring production AI agents
Development and testing
Reasonable API usage
Standard data retention

Not Allowed

Reselling credits
Mining or scraping data
Excessive API hammering
Storing non-agent data

⚠

Fair Use: Unusual usage patterns may be flagged for review. Contact support if you have questions about your use case.

Credit FAQs

Do credits expire?

Monthly plan credits reset each billing cycle. Enterprise plans can have custom terms.

Can I transfer credits?

No, credits are non-transferable between accounts.

What happens if I run out?

Your data collection pauses until credits are added or the billing cycle resets. Historical data remains accessible.

Can I get a refund?

Monthly subscription fees are non-refundable. Unused purchased credits can be refunded within 30 days.

How do I estimate my needs?

Start with the Scale plan and monitor your usage for a month. We'll help you optimize or upgrade as needed.

Monitoring Best Practices

Set Budgets

Create monthly credit budgets and threshold alerts (for example, notify Slack at 70% and email at 90%) from the dashboard at https://app.agenticants.ai. Budgets are a dashboard feature; there is no budgets API in the SDK.

Review Weekly

Check your credit usage every week in the dashboard at https://app.agenticants.ai. The Usage view compares the current week against the previous week and shows the percentage change. Weekly usage reporting is a dashboard feature, not an SDK call.

Optimize Continuously

Review optimization recommendations in the dashboard at https://app.agenticants.ai, which highlights where you can cut credit usage and the estimated monthly savings for each action. Optimization recommendations are surfaced in the dashboard rather than through the SDK.

Next Steps

Learn about FinOps - Optimize AI costs
Understand Pricing - See all pricing plans
Set Up Budgets - Control spending

Explore FinOps →