THE VINYL ACADEMYThe Meta Course For Meta AI
The Meta Course · seven chapters · one sitting

Use Meta AI like you mean it.

Everything a working professional needs to actually use Meta AI and Llama — including the parts that changed while everyone else's course stayed the same.

Two separate product lines, one confusing name. Know which is which, what it costs, what the license really says, and who can read your conversation.

One training session, start to finishEvery claim checked against MetaCheckpoints after every chapter
Start the course →
One name, two separate products
Meta AI

The assistant inside WhatsApp, Instagram, Messenger and Facebook. It runs on Muse Spark, the family Meta announced in April 2026 — closed weights you cannot download.

Llama

Models you download and run on your own hardware. Meta's position: unless you send inputs or outputs to Meta or publish them, Meta "cannot access, view, or use" them.

Meta AI no longer runs on Llama. The two lines answer the privacy question in opposite ways, which is why chapters six and seven treat them separately.

1

What "Meta AI" actually is now

By the end of this chapter: you can find Meta AI on every app you already have, and you can explain the one change that makes most Meta AI training obsolete.

Meta AI is the assistant Meta puts inside apps you already open every day — WhatsApp, Instagram, Messenger, Facebook — plus a standalone app, a website, and their glasses. You don't sign up for anything: Meta says "no separate account is needed beyond your existing Meta account." That's the whole pitch, and it's a real one. It's already on the phone in your pocket.

Start here, because almost every other Meta AI course is now wrong. Meta AI no longer runs on Llama. As of April 2026 the assistant is powered by Muse Spark, a new model family from Meta Superintelligence Labs. Llama and Meta AI are now two separate product lines. If you hear someone say "Meta AI is powered by Llama," they're describing last year's product. Chapters six and seven cover Llama on its own terms — because it's still very much alive, just no longer the thing answering you in WhatsApp.

Where it is, and how you actually start it

Same assistant, six front doors. The invocation differs in each one, which is the bit people fumble in front of a room.

Six ways in
  1. The Meta AI app — at the bottom, tap Ask Meta AI and type. This is the fullest version.
  2. On the web — go to meta.ai and sign in with your existing Meta account.
  3. WhatsApp — tap the Meta AI icon, or just start a new chat with Meta AI.
  4. Instagram — tap the Meta AI icon in your DMs, or use it right from search.
  5. Messenger — open the Meta AI tab, or Chats → Start chat → AI chat.
  6. Glasses — on Ray-Ban Meta and Oakley Meta, say "Hey Meta," or tap and hold the touchpad.

In a group chat, you summon it by name. Type @Meta AI in a WhatsApp or Messenger group and it joins that message. Useful — and it comes with a catch big enough that it gets its own box below.

The one everybody should hear before they use it at work. WhatsApp and Messenger are end-to-end encrypted — but messages that mention @Meta AI, and messages from Meta AI, are not encrypted. Bringing the assistant into a private group chat takes that part of the conversation outside the encryption. That is not a reason never to use it. It's a reason to never summon it into a thread carrying anything confidential.
Checkpoint — before you move on
  • You've opened Meta AI in at least two different places and seen it's the same assistant.
  • You can explain that Meta AI runs on Muse Spark, not Llama — and why that matters.
  • You know how to summon it in a group chat with @Meta AI.
  • You can state the encryption caveat in your own words.
Next chapterWhat It CostsKeep going →
2

Free — until it isn't

By the end of this chapter: you can give an honest answer to "is Meta AI free?" — which is no longer a one-word answer.

For years the answer was simply "yes." That changed. Meta now has paid subscription tiers and has begun capping the heavier features. Here's their own wording, because the nuance is the whole answer: "Meta AI is free to use for everyday use. We've started testing usage limits on some of our more compute intensive features. When people hit their daily / monthly limit, they'll have the option to purchase a subscription to unlock more usage before their free limit resets."

Still true

Everyday use is free

Asking questions, drafting, everyday image creation — no subscription, no credit card. For most people, most days, the free tier is the whole product.

The change

The heavy stuff is capped

The compute-intensive features — lots of image and video generation, and the deeper reasoning mode — now have daily and monthly limits. Hit one and you're offered an upgrade or a wait.

The paid product

It's called Meta One

Personal tiers are Meta One Plus and Meta One Premium. There are separate business and creator tiers: Essential, Advanced, Expert and Max.

What Premium adds

More, and deeper

Meta's list: generate more images and videos with expanded access to advanced models, more advanced reasoning via Thinking mode, and a larger monthly AI usage allowance.

On price — here's the honest position, and it's worth teaching as-is. Meta does not publish a price for any Meta One tier on any of their own pages. You'll find figures floating around online; they trace back to third-party posts, not to Meta. So the accurate statement is: pricing varies by tier and region, and Meta shows it to you in the app. If someone quotes you a confident number from a blog, that's a number nobody verified. Check it in-app, in your country, on the day.
Availability

Not everywhere yet

Meta describes Meta One as in "limited testing" and says it "isn't available everywhere yet." It's mobile only — iPhone or Android — and requires a personal, not business, Meta account. Don't promise it to a room that can't get it.

The billing detail

Nothing rolls over

Straight from Meta: "Unused usage for AI benefits from a previous billing cycle does not roll over." Use it or lose it each cycle — so buy the tier that matches a normal month, not your busiest one.

Checkpoint — before you move on
  • You can answer "is Meta AI free?" in a full sentence rather than one word.
  • You know the paid product is called Meta One, and can name the two personal tiers.
  • You know why you should never quote a Meta One price from an article.
  • You know unused allowance doesn't roll over.
Next chapterTalking To ItKeep going →
3

Talking to it — modes, voice and camera

By the end of this chapter: you've used all three modes on purpose, and you've held a real voice conversation instead of typing.

Meta AI's biggest advantage isn't raw capability — it's that it's already on the phone in your hand, and it's genuinely good at being talked to rather than typed at. That changes when you use it: in the car, on the shop floor, walking the yard.

Three modes — pick one on purpose

In the Meta AI app, tap the mode name in the bottom-right corner to switch. Most people never notice it's there and stay on the default forever.

Instant

Quick answers

The fast lane. Short questions, quick drafts, "what's the conversion for…" — you want the answer now, not a considered essay.

Thinking

The considered one

Meta's deeper reasoning mode, for complex questions and multi-step problems. It's also one of the features that counts against the paid allowance — so use it when the question earns it.

Shopping

Finding and comparing

Tuned for products — comparing options, finding something specific. Meta AI can also pull in recommendations from Reels and maps as you talk.

Voice — where it genuinely pulls ahead

Meta's description of what Muse Spark unlocked: you can "interrupt, switch topics, or swap languages" mid-conversation. That's much closer to talking to a person than to operating a device.

In the app

Tap the microphone

Tap the microphone icon before sending, and talk. You don't have to wait for it to finish — interrupt it, the way you would a colleague who's going off track.

Try"Walk me through what I should ask a freight broker — actually, hang on, make it for LTL specifically."
Point the camera

Ask about what you're seeing

During a voice conversation, tap the camera icon to ask about your surroundings — a label, a part, a form, a room. This came to glasses first and reached the app in 2026.

On glasses

Hands stay busy

Say "Hey Meta" or tap and hold the touchpad. Follow-up questions don't need the wake word again — so it's a conversation, not a series of commands.

Two things to know

The edges

There's an opt-in full-duplex demo under Settings → Voice, app only. And a plain limitation worth remembering before you demo it: voice conversations can't generate images.

Checkpoint — before you move on
  • You've switched between Instant, Thinking and Shopping on purpose.
  • You know Thinking mode draws on the paid allowance.
  • You've held one voice conversation and interrupted it mid-answer.
  • You've pointed the camera at something real and asked about it.
Next chapterImagesKeep going →
4

Images — the part that got seriously good

By the end of this chapter: you've made and edited a real image, and you know what's invisibly attached to every one you create.

Image generation is where Meta AI stopped being a novelty. It runs on Muse Image, and the interesting part is how it works — Meta's own description: it "doesn't just build an image; it thinks through your prompt first," planning the layout, looking up real-time web context, and blending multiple references before it draws anything.

What it's good at

The things that used to fail

Legible text inside the image — the classic AI weak spot — plus working QR codes, precise object removal, and blending several reference images into one composition.

How you edit

Mark it up

You don't have to re-describe the whole picture. Sketch or mark on the image to show what you want changed, or work from the presets panel. Point at the problem instead of explaining it.

Try"Remove the cable on the left and put our name on the crate in the same style as the label."
A genuinely useful one

Room redesign, shoppable

Redesign a room and the results can be made shoppable against Facebook Marketplace — the generated look connects to things you can actually buy. That's a real workflow, not a demo.

The cost side

Everyday free, heavy is capped

Meta's newsroom wording: free for everyday creation, and for people who want to create more, it's part of the subscription plans. Volume is exactly where the caps bite.

Every image you make is watermarked — invisibly. Meta attaches Content Seal, a hidden provenance signal that stays in the file "even when cropped, compressed, resized, or screenshotted." There's a detection tool at meta.ai/identification. Teach this as a feature, not a warning: it means an AI-generated image stays identifiable as one long after it leaves your hands. If you're putting AI images into customer-facing work, that's a fact you want to know before someone else discovers it.
Checkpoint — before you move on
  • You've generated one image with readable text in it.
  • You've edited an image by marking it up rather than re-describing it.
  • You can explain Content Seal and what survives a screenshot.
  • You know heavy image generation is the thing most likely to hit a cap.
Next chapterFor BusinessKeep going →
5

For business — the agent that answers your customers

By the end of this chapter: you know what Meta actually offers a business, what's generally available versus waitlisted, and the one disclosure you owe your customers.

This is the chapter that matters if you sell anything. Meta's business play isn't a chatbot on your website — it's an agent living inside the messaging apps your customers already use to contact you. Meta says "more than one million businesses" are already running one on WhatsApp and Messenger.

What it does

It works the whole conversation

Meta's list: answer questions specific to your business, make product recommendations from your catalog, book appointments, qualify incoming leads, let a human step in when you choose — and close sales.

What it costs today

Free to start — for now

Meta: businesses can activate an agent and "getting started is free." They also say that "in the coming months" it moves to paid subscriptions with options for businesses of every size. Set it up now, budget for later.

Where it runs

WhatsApp, Messenger, Instagram

The customer-response agent is being expanded to businesses of all sizes globally — that's the part you can actually count on today.

What's still waitlisted

The assistant-to-the-owner

The other half — morning briefings and running your operation for you — is limited to "a select number of businesses" with a waitlist. Don't build a plan on this half yet.

The two other business tools — and their real status

Meta ships a lot and announces earlier than it ships. Here's what's actually usable versus what's still a headline.

1
Business Agent Platform

Connects your agent to hundreds of systems — Shopify, Zendesk, Shopee — so it can act on your behalf, with enterprise-grade controls.

2
Meta AI Business Assistant — beta

Help with ads and account questions inside Business Suite, Ads Manager and Instagram Boost. Still beta, English-first, rolling out through 2026.

3
Muse Image in Advantage+ — not yet

Meta says advertisers will tap Muse Image through Advantage+ creative "in the coming weeks." Announced, not shipped. Check before you promise it.

The disclosure you owe your customers — read Meta's own words. On business chats using Meta AI: "These chats will be visible to you, the business and Meta." And: "Your feedback and messages may be used to improve AI at Meta." So a customer messaging your business is talking to you and to Meta. That's not a scandal — it's the deal. But it means you don't route account numbers, payment details or anything sensitive through an AI-handled thread, and you make sure a human takes over before the conversation gets to that point.
Checkpoint — before you move on
  • You can name three things a Business Agent does for a customer.
  • You know it's free to start and that Meta has said paid tiers are coming.
  • You can tell the difference between the GA customer-response agent and the waitlisted owner-assistant.
  • You can state who can see a business chat — and what that means for sensitive details.
Next chapterLlamaKeep going →
6

Llama — the other product line entirely

By the end of this chapter: you can explain the split between Meta AI and Llama, name the current models, and say honestly whether running one yourself makes sense for you.

Now the part that separates people who know this subject from people who read about it. Llama is not Meta AI. Llama is a family of models you can download and run on your own hardware — the reason Meta matters to businesses that can't send their data anywhere. Meta AI is the consumer assistant, and it runs on Muse. Two lines, one company, completely different jobs.

Meta AI / Muse

You use it. You can't have it.

Muse Spark and Muse Image power the assistant. They are closed weights — available by API to select partners in private preview. Meta says they "hope to open-source future versions," which is a hope, not a commitment.

Llama

You download it and own the runtime

Llama models you can run on your own machines. Nothing leaves your building. This is the entire reason Llama exists in a business conversation.

The killer fact

Meta can't see it — on one condition

Meta's own answer, verbatim: "Unless you provide inputs or outputs directly to Meta, or publish them, Meta cannot access, view, or use your inputs or outputs." For regulated or confidential work, that sentence is the whole pitch.

Where it lives now

The address changed

Heads-up if you've bookmarked anything: llama.com now redirects — the models and docs live at developer.meta.com/ai. Old links in old courses simply bounce.

The current lineup

Two generations are live. The numbers matter less than the shape: one family is built for scale, one for running small and cheap.

1
Llama 4 Scout

109B total parameters, and a class-leading 10 million token context window. Meta markets it as fitting single-GPU efficiency — the realistic starting point for a business.

2
Llama 4 Maverick

400B total parameters — Meta's most powerful open multimodal model. Not a single-GPU proposition; this one needs real infrastructure.

3
Llama 3.3 70B

The multilingual workhorse, and Meta's stated replacement for Llama 3.1 70B, which has been retired from the download page.

4
Llama 3.2 — 1B and 3B

The lightweight pair, cheap enough to run on mobile and edge devices. A model on a handheld scanner is a real thing you can do.

5
Llama 3.2 — 11B and 90B

Open multimodal models that reason over high-resolution images and return text. Photograph a damaged pallet; get a written condition note.

6
Safety models ship alongside

Llama 4 comes with Llama Guard 4 and Prompt Guard 2. Language support differs by generation: 3.1, 3.2 and 3.3 cover 8 languages; Llama 4 covers 12.

Getting one — and the honest hardware answer

Downloading is free and gated by a form. Running it well is where the real cost lives, and nobody should sell you past that.

From nothing to running
  1. Request access at developer.meta.com/ai. The form wants your legal name, date of birth, email, country, organization and job title — Meta warns that acronyms or missing corporate identifiers can block your access.
  2. Or skip the form — the models are also officially distributed via Hugging Face and GitHub.
  3. Or don't host at all. AWS, Azure, Google Cloud, Oracle, Databricks, Snowflake, Groq and others host Llama for you. Note their own terms apply on top of Meta's.
  4. Match the model to your hardware. Start at Scout or the small 3.2 models. Maverick at 400B parameters is a data-centre conversation, not a server-closet one.

If you want Muse instead of Llama — Meta offers the Meta Model API, currently in public preview for US developers only, and OpenAI-SDK compatible. Published rates: $1.25 per million input tokens, $0.15 cached input, $4.25 output, and $2.50 per 1,000 web-search-grounded queries, with $20 in free credits to start. Rates move — check before you budget.

Checkpoint — before you move on
  • You can explain the Meta AI / Llama split to someone in one sentence.
  • You can name the two Llama 4 models and say which one a normal business starts with.
  • You can quote the fact that Meta can't see what you run yourself.
  • You know llama.com now redirects to developer.meta.com/ai.
Next chapterLicense & RulesKeep going →
7

The license, and the rules

By the end of this chapter: you know what the Llama license actually permits, the one clause that stops European companies cold, and what happens to everything you type into Meta AI.

People call Llama "open source." It isn't — and Meta doesn't claim it is. They call it "a bespoke commercial license." The good news: for a normal business, it's genuinely permissive. The catch is that the parts people repeat at conferences are usually the parts they get wrong.

What you get

A broad, royalty-free grant

You may use, reproduce, distribute, copy, create derivative works of, and modify the models — worldwide, royalty-free. And you own your derivatives: as between you and Meta, modifications you make are yours.

Myth #1 — busted

You can train on the outputs

Everyone says you can't. That was true for Llama 2 and 3. From Llama 3.1 onward it's allowed, provided you include the correct attribution. If you heard otherwise, you heard about an old license.

Myth #2 — defused

The 700-million-user clause

Cited constantly as a blocker. Read it properly: it applies only if, on the model's release date, you had over 700 million monthly active users. It's a snapshot test aimed at Meta's rivals. For a normal business it is simply irrelevant.

What you owe

Say "Built with Llama"

You must prominently display "Built with Llama," ship a copy of the license and a NOTICE file, and if you fine-tune and distribute a model, its name must start with "Llama."

What "prominently" means — Meta gives a worked example, and it's a good one. If a developer released a four-page article about a new AI tool and mentioned "Built with Llama" only on the last page or in a footnote, Meta says they would not consider that a prominent display. So it isn't a legal-footer job. Put it where a reader will actually see it.

The clause that stops European companies

This is the single most business-relevant term in the whole license, and it's missing from almost every article written about it.

EU companies are not granted rights to the multimodal Llama models. Meta's wording: the rights "are not being granted to you if you are an individual domiciled in, or a company with a principal place of business in, the European Union." It applies to the multimodal models in Llama 3.2, 3.3 and 4. Three clarifications that matter in practice: end users of a product built on them are not affected; a non-EU company may still distribute into the EU and may employ EU-based developers; but an EU-domiciled individual can't use them personally. If you have European operations, this is the slide your legal team will want.

What you may not use it for

The Acceptable Use Policy is part of the license, so breaking it breaks your rights. Most of it is the obvious list — but three items catch ordinary businesses by surprise.

Surprise #1

Professional advice

No unlicensed practice of a profession — Meta names financial, legal and medical/health specifically. An AI that answers customers' financial or medical questions is squarely in this territory.

Surprise #2

Critical infrastructure and heavy machinery

Not for the operation of critical infrastructure, transportation technologies, or heavy machinery — and Meta is explicit that this holds "regardless of government certifications."

Surprise #3

Don't pass it off as human

You may not represent that outputs are human-generated, and you must disclose known dangers of your AI system to end users. That's a marketing-copy rule as much as a legal one.

The rest

The expected list

Illegal activity, harassment, unlawful discrimination in hiring/credit/housing, malware, weapons, disinformation, impersonation and fake reviews. If you need an exception, Meta reviews requests case by case.

What happens to what you type

Different question, same importance. And the answer differs sharply depending on whether you're using the assistant or running a model yourself.

Meta AI

It's retained and used

Meta's own words: "When information is shared with AIs, the AIs will sometimes retain and use that information," and "Do not share information that you don't want the AIs to use and retain." Meta also reviews conversations, automatically or manually.

The private option

Incognito Chat

Meta says of it: "no one — not even Meta — can read your conversations," conversations aren't saved, and messages disappear by default. Note it's still rolling out across WhatsApp and the Meta AI app, so don't assume you have it.

Region matters

EU and UK get an objection right

Users there receive notice and an objection form under GDPR to object to their public data being used to train Meta's models, along with access, deletion and correction rights. Elsewhere, no equivalent training opt-out is documented.

Self-hosted Llama

The clean answer

Run it yourself and none of the above applies: unless you send inputs or outputs to Meta or publish them, Meta states it cannot access, view or use them. If the data can't leave, that's your answer — and it's the strongest argument for Llama in a regulated business.

Take control of your own data — right now
  1. Open the Meta AI app and go to Menu → Settings.
  2. Open Data & privacy → Manage your information.
  3. From there you can export your information, remove public content, or delete all chats and media.
  4. Then set the rule you'll actually follow: sensitive material never goes into the assistant — not in a chat, and not by summoning it into a group thread.

Say the quiet part out loud. Outside the EU and UK, Meta documents no toggle that removes your conversations from training. Deletion and discipline are the controls you have — so the discipline is the real control.

Final checkpoint — the first-week five
  • Use Meta AI in the app you already live in, and try one voice conversation (ch. 1, 3).
  • Generate and edit one image, and know Content Seal rides along with it (ch. 4).
  • If you sell anything, look at a Business Agent — and decide where a human takes over (ch. 5).
  • Be able to explain the Meta AI / Llama split, and the EU multimodal restriction, without notes (ch. 6, 7).
  • Set your rule for what never goes into the assistant — and verify every fact, figure and name before it leaves your hands.