Skip to main content
LLM Metrix
Back to changelog
v0.0.16Aug 9, 2026Release

v0.0.16: Simpler Plans, Stronger Alerts & Brand Safety

The plan ladder became three plans with daily prompt allowances, brand-safety findings now run claim by claim with sentiment and risk, and alerts are tunable per trigger.

Plans you can explain

Three Plans, One Credit

A note on the plan ladder. This entry describes prices as they shipped then (Business $29.99/mo, Agency $199.99/mo). Today’s figures are on the pricing page. The sentence below reflects what shipped then.

The plan ladder is now three plans: Free, Business ($29.99/mo) and Agency ($199.99/mo), with a uniform 15% off for annual billing. Every plan has a per-period prompt allowance (3 prompts a week on Free, 15 a day on Business, 100 a day on Agency), and one credit is one prompt, checked once across the engines you run. Paid plans include any four of the six AI models, chosen by you; Agency can add the rest per engine (Grok and Meta AI are available today). Business covers 3 domains and 3 seats; Agency covers 100 domains and 10 seats.

Brand Safety, Claim by Claim

The brand-safety view now works claim by claim: each claim an engine makes about your brand is scored for accuracy risk and sentiment, and the positive / neutral / negative split is shown as a distribution over time. Every finding links to the specific remediation it points at, so a claim that hurts you names the fix, not just the problem.

Automated Alerts You Can Tune

Alerts are now per-trigger: each trigger type (position drops, share-of-voice shifts, citation losses) has its own threshold you can edit and its own frequency mode, so a low-stakes signal can go into a daily digest while a position collapse pings immediately. See the setting up alerts tutorial. A baseline-deviation mode alerts when your metrics drift from their own recent history, and a position-drift mode watches ranks shift across scans. Delivery is by email or Microsoft Teams webhook.

Bulk-Import Your Tracked Prompts

Tracked prompts can be imported in bulk from CSV or JSON, with validation reporting exactly which rows were rejected and why.

A Grounded llms.txt Generator

The llms.txt generator now reads your sitemap, grounds the generated file in your pages, validates the result, and can emit llms-full.txt for sites that want the fuller view.

Citations, Per Market

The citations view gained a market selector, so citation sources can be examined for a specific region rather than mixed together.

A Bigger, Better Shell

The navbar menu was rebuilt as a viewport-aligned surface with featured rails and live filtering, onboarding shows the six engines you can choose from in a preview carousel, and the dashboard shell tightened up: refined domain context controls, a responsive layout, and no white flash on scroll.

Want to shape what we build next?

See what we're building next, and tell us what you need.