Pricing & Usage

This document explains how costs are calculated per scan build, how tokens are tracked, and what LLM models are available for use.

---

Cost Components

Every build accumulates costs across four distinct categories. The total cost for a build is the sum of all four:

Total Cost = LLM Cost + Cloudflare Workers Cost + R2 Storage Cost + Browser Cost

---

1. LLM Cost

The LLM cost covers AI inference — the tokens consumed when the scanner sends prompts to and receives responses from the chosen language model.

LLM cost is broken into up to four sub-components:

Sub-component	Formula
Input cost	(prompt_tokens / 1,000,000) × input_rate
Output cost	(completion_tokens / 1,000,000) × output_rate
Cache write cost	(cache_write_tokens / 1,000,000) × cache_write_rate (Claude models only)
Cache read cost	(cache_read_tokens / 1,000,000) × cache_read_rate (Claude models only)

LLM Cost = Input Cost + Output Cost + Cache Write Cost + Cache Read Cost

All rates are per 1 million tokens and vary by model (see Supported Models & Pricing below).

Prompt caching (cache write / cache read) is only available on Anthropic Claude models. For all other models, cache costs are $0.

---

2. Cloudflare Workers Cost

The Cloudflare Workers cost covers CPU execution time used by the scanner worker during a build.

Rate	Formula
$0.02 per 1M CPU milliseconds	(cf_workers_duration_ms × 0.02) / 1,000,000

cf_workers_duration_ms is the total CPU wall-clock time the Cloudflare Worker was active during the build.

---

3. R2 Storage Cost

The R2 cost covers object storage operations and storage space for scan artifacts (raw findings, reports, etc.).

Component	Rate	Estimate
Class A ops (PUT/writes)	$0.36 per 1M ops	~(api_calls × 2 + 3) ops per build
Storage	$0.015 per GB/month	~25 MB per build

r2_ops  = (api_calls × 2) + 3
r2_cost = (r2_ops / 1,000,000) × 0.36 + (25 / 1024) × 0.015

The + 3 accounts for fixed operations per build: ingestion, normalization write, and report generation.

---

4. Browser Cost

Browser cost applies only when Cloudflare Browser Rendering is used (e.g., for DAST / web scanning). It has two parts:

Browser Runtime

Component	Rate
Session duration	$0.09 per hour
Concurrent browser (daily amortized)	$0.066 per day

cf_browser_run_cost = (browser_duration_sec / 3600) × 0.09 + 0.066

Browser LLM

The same input/output rates as the main LLM apply to tokens consumed during browser-assisted AI analysis.

browser_llm_cost = (browser_prompt_tokens / 1M) × input_rate
                 + (browser_completion_tokens / 1M) × output_rate

Total Browser Cost = Browser Runtime Cost + Browser LLM Cost

If the browser was not used in a build, all browser costs are $0.

---

Token Tracking

Each build records the following token fields:

Field	Description
prompt_tokens	Tokens sent to the model (input / prompt)
completion_tokens	Tokens generated by the model (output)
total_tokens	Sum of prompt + completion tokens
cache_write_tokens	Tokens written to the prompt cache (Claude only)
cache_read_tokens	Tokens read from the prompt cache (Claude only)
api_calls	Number of LLM API calls made during the build

Token counts are reported by the LLM API and aggregated across all calls in a single build. Monthly totals are accumulated per organization in the usage_monthly collection.

---

Credits & Limits

Usage limits are enforced at the organization level per calendar month.

AI Credits (Builds)

Setting	Default
Builds per month	20

Each completed build consumes 1 AI credit. If your organization has exhausted its monthly credit limit, new enrichment runs will be blocked until the next billing cycle or until additional credits are purchased.

Credit ledger value is stored as:

total_credits = llm_total_cost × 2.5

Runtime Minutes

Setting	Default	Overage Rate
Runtime per month	3,000 minutes	$0.05 / minute

runtime_duration_ms tracks cumulative Cloudflare Worker CPU time across all builds in the month. Exceeding the runtime limit blocks new scanner runs.

---

Monthly Aggregation

Per organization, the following fields are aggregated in the usage_monthly collection (reset each calendar month):

Field	What it accumulates
total_tokens	All prompt + completion tokens across builds
total_llm_cost	LLM cost (main + browser LLM)
total_cf_cost	CF Workers + R2 + CF Browser runtime
total_cost	Grand total across all cost components
total_credits	Credit ledger (LLM cost × 2.5)
runtime_duration_ms	Total CF Workers CPU time
total_builds	Number of completed builds

---

Supported Models & Pricing

All prices are per 1 million tokens (USD).

OpenAI

GPT-5 Family

Model	Input	Output
gpt-5	$0.625	$5.00
gpt-5-mini	$0.25	$2.00
gpt-5-nano	$0.05	$0.40

GPT-5.x Family

Model	Input	Output
gpt-5.1	$1.25	$10.00
gpt-5.2	$1.75	$14.00
gpt-5.4	$2.50	$15.00
gpt-5.4-mini	$0.75	$4.50
gpt-5.4-nano	$0.20	$1.25

GPT-4.1 Family

Model	Input	Output
gpt-4.1	$2.00	$8.00
gpt-4.1-mini	$0.40	$1.60
gpt-4.1-nano	$0.05	$0.20

GPT-4o Family

Model	Input	Output
gpt-4o	$2.50	$10.00
gpt-4o-mini	$0.15	$0.60