Jev is the new kid on the block. Everyone is currently on the hype train. But I wanted to give it a proper whirl and see if it lives up to the hype, rather than hopping on the bandwagon saying "iTs A GaMeChAnGer". If you don't know what Jev is, it's a new type of AI model. It's not an LLM that predicts text token-by-token (like what you're used to). It's an AI model based on probability. You give it a set criteria -> you give it data to ingest -> it determines the "highest" probability based on the criteria you've provided -> gives you an output. Think of it like a 100x faster Sonnet or GPT-Luna. Except it can't think. It's not a reasoning model. It's literally based on probability. It won't be able to answer: "is this a cat or a dog". It WILL be able to answer "out of 10, how spammy is this email" (given that you've provided it a criteria). This is the best visual explanation of the difference between Jev & LLMs: 👉 https://x.com/MatijaSosic/status/2100190746389135772/video/1 People have used it to: - Classify emails and route them to the right person - Qualify leads & route them to the right sales rep - Create a "AI Slop" filter on LinkedIn And a bunch of other cool use cases. The best thing about this model is that: - Input tokens: $0.0421 / 1M - Output tokens: $0.00 (FREE) IT'S AS CHEAP AS CHIPS! So I've thought about how we could apply this to SEO/GEO. There are some absolutely awesome use cases I'm keen on testing. So over the next couple of days, I'll be building out the following SEO/GEO skills: 1️⃣ Internal Links: Scrape your website, and get recommended internal links to add to achieve better topical authority. 2️⃣ Topic Clustering: Cluster your keywords into "topics" as efficiently as possible. 3️⃣ GSC Keyword Classifying: Classify your "keyword" list in GSC by buyer's journey, so you can determine which keywords are BoFu, MoFu or ToFu. 4️⃣ Content Audit: Determine whether you should keep, update or purge old content (on-site).