Verging Memory CI billing

This page explains the standard plan arithmetic. Your Order controls if it states different prices, included usage, weights or terms. The Terms of Service govern charges, payment, disputes, adjustments and voiding.

Standard plan

Unless an Order states otherwise, the plan costs USD $22 per billing period. The plan includes 40 environments. Overage costs USD $0.715 per environment, including Activation usage.

Overage accrues on your balance and is charged to the saved payment method when the balance reaches USD $22, at renewal, or when cancellation or termination takes effect. This automatic charge is a top-up and is on by default.

Calculate environments

Regression reports show Environments tested, which counts each test suite exercised in each agent setup once. Plan accounting separately uses Environment usage, which applies the weights below and is the amount deducted from your plan.

For ordinary release testing, each tested Environment contributes:

test suite weight × release agent setup weight

Activation applies to the first delivered testing of each test suite in each agent setup. For that combination, use the activation agent setup weight instead. Storage-only testing is available only for ordinary releases and uses half the otherwise applicable amount. Add the selected suite and setup amounts, then round the release total up to the next whole environment.

When one release includes both suites already set up on their agent setups and new suite-and-setup combinations being activated by Verging Labs, plan accounting records two lines: ordinary Release usage for the existing combinations and Activation usage for the new combinations. Onboarding does not require Activation.

Test suite weights:

• Core Recall: 1
• Preference Adherence: 1
• Onboarding: 1
• Truth Maintenance: 2
• History Migration: 2
• Memory at Scale: 3
• Long-Horizon Retention: 10, available on a separately ordered tier

Release agent setup weights:

• Hermes GPT-5.6 Luna: 1
• Hermes GPT-5.6 Terra: 2
• Hermes GPT-5.6 Sol: 4
• Claude Code Opus 5: 4

Default activation agent setup weights:

• Hermes GPT-5.6 Luna: 2
• Hermes GPT-5.6 Terra: 2
• Hermes GPT-5.6 Sol: 4
• Claude Code Opus 5: 4

Activation overrides:

• Memory at Scale on Hermes GPT-5.6 Sol or Claude Code Opus 5: 7
• Long-Horizon Retention on Hermes GPT-5.6 Sol or Claude Code Opus 5: 8

More test questions

If your onboarding details include a 2× or 4× question expansion, Verging Labs asks two or four times the standard number of questions for the named suites. Those suites use two or four times their ordinary environment usage in every selected agent setup. Question expansion is configured on your account during onboarding; it is not a workflow or API request field.

Retrieval budget

Each Environment includes a budget for the number of tokens your memory tool returns to the agent. The budget is the question count for that suite multiplied by a per-question amount:

• Core Recall, Preference Adherence, Truth Maintenance, History Migration: 2,500 returned tokens per question
• Memory at Scale, Long-Horizon Retention: 5,000 returned tokens per question

Only what your memory tool hands back counts toward the budget. Verging Labs' prompt and the agent's own reasoning do not.

If a suite returns more than its budget, the extra usage is charged proportionally. The Environment's usage increases by its weighted amount multiplied by the fraction it went over budget, and the release total is still rounded up to the next whole environment. Going over budget is a billing condition, not a failed test, and it does not change any test result. Your report names any suite that went over its budget.

Payment issues

If a payment fails or requires additional authentication, Verging Labs may hold new releases until payment clears. Reports already delivered remain available.

For the complete contractual rules, including billing periods, taxes, cancellation, disputes, failed payments and allocation when testing fails, read the Terms of Service.

See Integrate Verging Memory CI with GitHub Actions, API setup, and Agent setups.