New engines, free tools and a sharper product - every change that lands in Zene, newest first.
The new Chrome extension checks any page's AI readiness without leaving the tab - open a competitor's post, your own article or a draft, and get an instant verdict on whether AI assistants can quote it, plus the specific fixes to make it quotable. There's a dedicated /extension home for it on the site.
Paste any URL and Zene runs a real, multi-engine scan on the spot - see whether ChatGPT & friends surface the brand, then drop into a full GEO audit of the page: meta, headings, structure and the answer the AI actually gives. Compare two URLs side by side and sweep a whole sitemap. The old standalone audit and readiness tools now live here as one tool.
A visual reset across the app and site: squared-off edges, crisper lines and flatter surfaces replace the soft rounded look, with tighter engine cards, animated metric nodes and a brand marquee. Less decoration, more signal.
xAI's Grok joins ChatGPT, Claude, Gemini and Perplexity as a tracked engine, so your visibility now spans the five assistants people actually ask. Grok scanning is available on Pro and Agency, with its own engine card and accent throughout.
Signup got smarter about your market: Zene discovers your competitors and the topics you compete on automatically - web-grounded and location-aware, so a London brand sees London rivals. The reveal ends on a head-to-head leaderboard, and you can pick up right where you left off if you bounce.
Your visibility score and the two fast engines (ChatGPT & Gemini) still refresh every day. The two heavier engines (Claude & Perplexity) now refresh on a multi-day cycle instead of daily - weekly on Pro, every 5 days on Agency. Plainly: premium answers update less often by default. In exchange you get three things - a Re-scan button that pulls fresh premium data for any brand the instant you want it, premium engines that stay on every paid plan (not just the priciest), and faster, steadier scans overall. Each engine card now shows when its answer was last refreshed.
Two upgrades to how we measure you. Precision: brand detection now uses whole-word matching and checks every name against the model's actual answer, so "Notion" no longer counts a mention of "Notional" - a few brands will see a small, permanent score correction as the noise clears (the number is now truer, not lower). Stability: your headline score is now a 7-day moving average and alerts only fire when a change holds across two scans, so day-to-day model randomness stops jolting the number. Daily-plan brands get the steadiest scores.
Articles now do things: type your brand into the inline audit and it stages real engine questions on the spot, a quotable-sentence builder writes your positioning line, and an llms.txt generator ships you the file. Also: this page's time scrubber now travels in the right direction.
A new Zene mark - a chunky lowercase z on a pastel tile - across the favicon, the comparison pages and the footer. Blog covers swapped their washed-out pastels for rich duotone gradients, so the cover artwork finally reads at a glance.
Honest, dated side-by-side comparisons against Peec, Profound, Otterly and friends - real trade-offs and a "choose them if" section for every rival. Underneath: a full SEO/GEO foundation with sitemap, OG images, JSON-LD and llms.txt.
The dashboard's last mock is gone: streaks and the new /achievements page now compute from your live scans, actions and score history. Small dopamine, real data.
A live scan progress hero shows every engine as it answers, with real failure states instead of silent zeros. Cadence is now per-plan - daily on Pro and Agency - and an autonomous queue drainer makes sure first scans never stall.
Google One Tap lands on the public site and login. Your Google profile photo follows you in - OAuth avatars now persist to your Zene profile automatically.
Landing, features, blog and this changelog - rebuilt as a fast, GSAP-animated marketing site with scroll reveals and live count-ups, fully separated from the app shell.
New name, new home: tryzene.com. Same product, sharper identity. Billing moved to monthly & yearly plans in USD via Polar, and the frontend now ships as a standalone Docker image.
The signup flow now runs a real visibility probe: pick your prompts and we ask ChatGPT live - your favicon, an auto-written description and your first real answer appear before you've even created a password.
Upload and crop brand logos and profile photos, add brands in a guided multi-step modal, and manage your whole account from Settings. Plan entitlements - brands, prompts, scans, AI assists - are now enforced end-to-end from a single source of truth.
Every answer now carries its citations. The new Sources screen shows which domains each engine trusts in your category, where you're absent - and turns each source gap into an action.
Actions are now an entity-type playbook - a plain, stateful checklist tuned to your brand type. Complete one and Zene re-scans automatically, measuring the real before/after score lift instead of promising one.
Scanning went multi-engine: ChatGPT, Claude, Gemini and Perplexity behind one engine registry, with versioned prompt templates, canonical query dedup and hard cost caps.
Supabase auth, Postgres with row-level security and a server-only data layer under every screen - the mock demo became a real product in one (long) day. A super-admin console ships alongside.