Cost Tracking

The AI Platform tracks the exact USD cost of every AI primitive consumed — tokens, voice minutes, phone calls, crawled pages, and governance scans. Cost tracking helps you understand your AI spend, attribute it to specific products and users, and forecast what next month will look like.

How Costs Are Calculated

Every AI call records the number of tokens used and the model that served the request. The platform applies per-token pricing to compute a USD cost:

Chat Token Pricing

Model	Input (per 1k tokens)	Output (per 1k tokens)
GPT-4.1	$0.002	$0.008
GPT-4.1-mini	$0.0004	$0.0016
GPT-4.1-nano	$0.0001	$0.0004
GPT-4o	$0.005	$0.015
GPT-4o-mini	$0.00015	$0.0006

Other Primitives Pricing

Primitive	Cost
Voice room	$0.005 per participant-minute
Phone call	$0.02 per minute
Phone SMS	Not currently metered
Web page crawl	$0.02 per page
Email send	$0.001 per email
Governance scan	$0.05 per scan
Safety moderation	$0.005 per request
Embedding generation	Charged at the input token rate for the embedding model

ℹ️These are the platform's internal metered costs — what the AI Platform charges against your monthly budget cap. Your actual subscription tier pricing is separate and does not scale with usage directly.

Viewing Cost Breakdown

Navigate to AI Platform → Usage Analytics → By Feature to see cost per feature.

Navigate to AI Platform → Usage Analytics → By Product to see cost attributed to each The One product.

Navigate to AI Platform → Usage Analytics → By User to see cost per team member.

Each view supports time period selection (7 days, 30 days, 3 months, 12 months, custom).

Monthly Cost Forecast

The Quota tab in Usage Analytics shows a projected month-end spend based on your current daily burn rate:

• Today's spend — Tokens and cost accumulated so far today
• This month's spend — Month-to-date total
• Projected month-end — Linear extrapolation of daily average to end of month
• Budget cap — Your monthly spend limit
• Days remaining — How many days until the budget resets

If the projected month-end exceeds your budget cap, a warning banner appears and you will receive an alert (if configured).

Cost by Feature — What Costs the Most

Common high-cost features and why:

Feature	Why it costs more	Mitigation
jarvis-assistant	GPT-4.1 model, long system prompt	Expected — Jarvis is the most capable route
attack-narrative	GPT-4.1 with long context (full process tree)	Only invoked on serious alerts — acceptable
proposal-generate	Long output (multi-page proposals)	Review output length settings in the route config
behavioral-analysis	GPT-4.1 with 30-day baseline context	Runs per-device on schedule — review frequency
chat-playground	User-selected GPT-4.1	Educate users on model selection for simple tasks

To reduce cost on a specific feature, consider routing it to a less expensive model in the feature routing config (Enterprise tier).

Budget Cap Management

Your monthly budget cap is the hard ceiling on AI spend for the month. Once the month-to-date cost hits the cap, new requests are blocked with HTTP 429 until the budget resets on the 1st.

To change your budget cap:

1. Navigate to Settings → Billing
2. Click Edit budget cap
3. Enter the new monthly limit in USD
4. Click Save

⚠️Lowering the budget cap takes effect immediately. If your current month-to-date spend already exceeds the new lower cap, requests will be blocked immediately.

Exporting Cost Reports

Export a CSV of your cost data:

1. Navigate to Usage Analytics
2. Select the tab and time period you want
3. Click Export → CSV

For accounting purposes, the CSV includes:

• Date
• Feature / primitive
• Provider
• Model
• Request count
• Token count (input and output separately)
• Cost (USD)

Comparing Cost to Subscription Pricing

The cost shown in the AI Platform is your AI usage cost, which is applied against your monthly budget cap. This is separate from your subscription tier fee (Starter $X/month, Professional $X/month).

Your invoice from The One Stack shows:

• Subscription tier fee
• Studio platform fee (if subscribed)
• Studio app fees (per deployed app)
• Studio client fees (per client tenant)

AI usage cost does not appear as a line item on your invoice — it is a guardrail, not a separate bill. If you stay within your budget cap, there is no overage charge. If you need more budget, upgrade your tier.

API Access

GET /api/usage/cost-summary?period=30d
Authorization: Bearer oneai_your_key_here

Response:

{
  "period": "30d",
  "total_cost_usd": 47.23,
  "by_feature": [
    { "feature": "jarvis-assistant", "cost_usd": 18.41, "requests": 1205 },
    { "feature": "ticket-suggest", "cost_usd": 12.87, "requests": 8943 },
    { "feature": "proposal-generate", "cost_usd": 8.12, "requests": 34 }
  ],
  "by_product": [
    { "product": "psa", "cost_usd": 15.22 },
    { "product": "ai-platform", "cost_usd": 23.18 },
    { "product": "defend", "cost_usd": 8.83 }
  ],
  "projected_month_end_usd": 52.10,
  "budget_cap_usd": 500
}