OverpayingForAIPricing desk

Meta · Beginner → Intermediate · 8 lessons · about 79 minutes · updated 2026-09-04

Llama 101: The Complete Beginner's Course (Pricing-First)

Open-weight models you can rent by the token or run yourself. Learn what that actually costs.

Beginners who have heard that Llama is "free" and want to know what that means in practice, developers choosing between a hosted API and self-hosting, and anyone comparing open-weight models against ChatGPT, Claude or Gemini on cost.

Meta models: hosted API price per 1M tokens (live from our catalogue)Live from the OverpayingForAI catalogue · last verified 2026-09-04 · sorted by output priceInputOutputMuse Spark 1.31.0M context$1.25$4.25Muse Glimmer 30B131K context$0.3$1.1Llama 4 Maverick1.0M context$0.2$0.696Llama 3.3 70B Instruct131K context$0.1$0.32Llama 4 Scout1.3M context$0.1$0.3Llama 3.1 8B Instruct131K context$0.05$0.08Batch, fast-mode and free variants are excluded. Prices change; the figure re-draws from the catalogue on every build.
Live Meta API prices from the OverpayingForAI catalogue. Lesson 3 explains how these relate to the subscription plans.

Course outline

  1. 1What is Llama?Llama is Meta's family of open-weight language models. This lesson explains what open weights mean, who Llama is for, and where it falls short.8 min · 3 objectives
  2. 2Llama models explainedScout, Maverick, 3.3 70B, 3.1 8B, plus Meta's Muse models: what each is for, what it costs per token on hosted APIs, and how much context you get.10 min · 3 objectives
  3. 3Pricing and plansThere is no Llama subscription. This lesson lays out the three real cost routes: the free Meta AI assistant, hosted APIs billed per token, and self-hosting.9 min · 3 objectives
  4. 4Getting started in 15 minutesTwo starting points: chat with Meta AI for free, or run a specific Llama model through a hosted API or a local runner. Both paths, step by step.12 min · 3 objectives
  5. 5Prompting basics for LlamaOpen-weight instruct models reward clear structure. Learn the anatomy of a good prompt and see three before-and-after examples that work on Llama.9 min · 3 objectives
  6. 6Everyday workflowsFive concrete workflows on Llama: writing, research, coding, data and planning. For each, the model to use and what it costs on Meta AI, a hosted API and self-host.10 min · 3 objectives
  7. 7API quickstartCall a Llama model through an OpenAI-compatible hosted endpoint with curl and Python, then run the same request against a local runner. Understand tokens before you pay for them.11 min · 3 objectives
  8. 8Cost controlThe cheapest way to get the result: route tasks to the smallest adequate model, trim context, cap output, know when self-hosting pays off, and spot the red flags of overpaying.10 min · 3 objectives

Frequently asked

What is the difference between Llama and Meta AI?

Llama is the family of open-weight models Meta publishes. Meta AI is the free consumer assistant inside WhatsApp, Instagram, Messenger and meta.ai that runs on Meta's models. You cannot pick a model in Meta AI; you can with a hosted API or a local runner.

How much does Llama cost?

There is no single price. Meta AI is free. Hosted APIs bill per token; at the time of writing our catalogue lists Llama 4 Scout at $0.10 per million input tokens on OpenRouter. Self-hosting costs hardware and electricity instead of tokens.

Which Llama model should a beginner start with?

Llama 4 Scout on a hosted API, or Llama 3.1 8B locally if you want to run it on a laptop. Move up to Llama 4 Maverick only when a task clearly needs it.

Is Llama open source?

Llama is open-weight under Meta's community licence, which is not a standard open-source licence. For most personal and small-business use this makes no practical difference, but read the licence if you are building a product.

Can I use Llama with the OpenAI SDK?

Yes. Most hosts and local runners expose an OpenAI-compatible endpoint. Set the base URL to your host or runner and use its model id.

Is it cheaper to self-host Llama or use an API?

For most people the API is cheaper because you pay only for tokens used. Self-hosting is a fixed cost that pays off at high volume or when data must stay private.

If our calculators helped you cut down on hidden AI wallet leaks, consider buying us a coffee. A tiny fraction of your savings keeps our pricing indexes updated daily.

Not sure which AI is cheapest for your use case? Find out in 30 seconds — no signup required.

AI cost intelligence

Stop overpaying for AI tools

Join the OverpayingForAI list for pricing updates, cheaper alternatives, and practical buying guidance.

Now tracking 50+ AI tools, models, platforms, subscriptions, coding tools, and automation products.

We use your email only for OverpayingForAI updates. Unsubscribe anytime.