OverpayingForAIPricing desk

Tutorials · the training desk

Learn AI from zero — then every major model, and exactly what it should cost you.

25 free courses, 140 lessons. Each one starts with what the model is, ends with the cheapest way to use it, and links straight to the comparisons and calculators on this site.

Start here

New to AI? Begin with these.

No vendor to choose yet. Learn what AI can do for you, how to ask, when not to trust it, and which tool fits your job. Every lesson has something to try and ends with whether you need to pay.

6 courses · 33 lessons

How a large language model answers youYou pay for what goes in (input tokens) and what comes out (output tokens). Nothing is remembered between calls unless it is sent again.Your prompttext, files, imagestokenisedInput tokenscounted & billedThe modelpredicts the next tokenone token at a timeOutput tokenscounted & billedOutput tokens usually cost 3–5× more than input tokens. Long answers, not long questions, are what run up an API bill.
Choose by need, not by brand1What do you mainly want to …write, documents, research, code2Start on a free tierall three big tools have one3Use it for two weeksnotice what stops you4Pay only for the thing that…$8 to $20 a month
Anatomy of a prompt that works first timeFive parts. Most bad answers are missing two of them.RoleWho the model should be
“You are a pricing analyst…”
TaskThe one thing you want
“Rank these 4 plans by cost per 1K tokens.”
ContextFacts it cannot guess
paste the plan table, your usage
FormatShape of the answer
“a table, then a 2-line verdict”
ConstraintsWhat to avoid
“no marketing language; cite the source”
Cost angleA precise prompt costs a few hundred tokens more and saves the two or three retries that cost thousands.Asking for a short format (“3 bullets”) cuts output tokens — the expensive kind — by half or more.
Vendor-neutral·Beginner·5 lessons · 28 min

How to Write a Good AI Prompt: Talk to AI Without Jargon

One three-part habit, five short lessons, no vocabulary to memorise. Get a usable answer on the first try more often, and know exactly what to say when you do not.

  1. 1Task, Context, Output: the only three parts a prompt has
  2. 2See it, copy it, change it
  3. 3Giving AI what it needs to know
  4. 4Asking for the shape you want
  5. 5Fixing a weak answer
Every win in this course follows the same loop1Paste the promptthe text is already insi…2Read the answertwo minutes, not twenty3Check itdid it do what you asked?4Adjust one thingshorter, warmer, more sp…5Decidefree tier fine, or worth…
Vendor-neutral·Beginner·5 lessons · 32 min

10 Things You Can Do With AI Today

Ten small, useful jobs — an email, a summary, a plan, a spreadsheet question — each one done in two minutes with a prompt you can paste, and a note on which ones a free account handles.

  1. 1Rewrite an email, then summarise something long
  2. 2Explain a difficult concept, then brainstorm ideas
  3. 3Turn notes into a document, then compare options
  4. 4Create a plan, then analyse a small file
  5. 5Prepare for a meeting, then learn something new
How a large language model answers youYou pay for what goes in (input tokens) and what comes out (output tokens). Nothing is remembered between calls unless it is sent again.Your prompttext, files, imagestokenisedInput tokenscounted & billedThe modelpredicts the next tokenone token at a timeOutput tokenscounted & billedOutput tokens usually cost 3–5× more than input tokens. Long answers, not long questions, are what run up an API bill.
The document loop1Upload or pastePDF, DOCX, or the text i…2Ask one thinga question, a length, a …3Read the answeragainst the document, no…4Check one figureopen the page it came fr…5Ask the next thingsame chat, same document
Vendor-neutral·Beginner·5 lessons · 33 min

How to Use ChatGPT, Claude and Gemini With PDFs (Free vs Paid)

Upload a document, ask it questions, summarise it to the length you want, pull out the action items, compare two versions, and know exactly where a free plan stops being enough.

  1. 1Upload a document and ask it three questions
  2. 2Summarise to a length you choose
  3. 3Extract information and action items
  4. 4Compare two documents and find what changed
  5. 5Turn documents into notes, and what free plans allow

One product, in depth

The 101 courses, by vendor.

Eight lessons each: what it is, the models, pricing and plans, the first 15 minutes, prompting, workflows, the API and cost control.

8 courses · 62 lessons

What one model costs per 1M input tokens$ per 1M input tokensQwen3.7 Flash0.03 bulk tierGPT-OSS 120B0.04 open weightsGemini 2.5 Flash Lite0.1 cheap + long contextDeepSeek V4 Flash0.14 cheap reasoningGPT-51.25 frontierClaude Sonnet 52 frontier
Claude API pricing — $ per 1M tokensLive from the OverpayingForAI catalogue · last verified 2026-10-02 · sorted by output priceInputOutputClaude Fable 5.11M context$10$50Claude Opus 4.81M context$5$25Claude Opus 51M context$5$25Claude Opus 5.51M context$4$20Claude Sonnet 4.61M context$3$15Claude Sonnet 51M context$2$10Batch, fast-mode and free variants are excluded. Prices change; the figure re-draws from the catalogue on every build.
OpenAI API prices, $ per 1M tokens (live from our catalogue)Live from the OverpayingForAI catalogue · last verified 2026-10-02 · sorted by output priceInputOutputGPT-5.51.1M context$5$30GPT-5.41.1M context$2.5$15GPT-5.6 Terra1.1M context$2$12GPT-5.6 Sol1.1M context$2$10GPT-4.11.0M context$2$8GPT-5.4 mini400K context$0.75$4.5GPT-5 Mini400K context$0.25$2GPT-4.1 Mini1.0M context$0.4$1.6Batch, fast-mode and free variants are excluded. Prices change; the figure re-draws from the catalogue on every build.
Gemini API prices, $ per 1M tokens (live from our catalogue)Live from the OverpayingForAI catalogue · last verified 2026-10-02 · sorted by output priceInputOutputGemini 3.1 Pro Preview1.0M context$2$12Gemini 2.5 Pro1.0M context$1.25$10Gemini 3.5 Flash1.0M context$1.5$9Gemini 3.6 Flash1.0M context$0.75$3.75Gemini 3.7 Flash1.0M context$0.75$3.75Gemini 3.8 Flash1.0M context$0.75$3.75Gemini 3 Flash Preview1.0M context$0.5$3Gemini 2.5 Flash1.0M context$0.3$2.5Batch, fast-mode and free variants are excluded. Prices change; the figure re-draws from the catalogue on every build.
DeepSeek API prices per 1M tokens (live from our catalogue)Live from the OverpayingForAI catalogue · last verified 2026-10-02 · sorted by output priceInputOutputDeepSeek R164K context$0.7$2.5R1 0528164K context$0.5$2.15DeepSeek V4 Pro 08131.0M context$0.66$1.98DeepSeek V4 Flash1.0M context$0.006$1.6DeepSeek V4 Flash 07311.0M context$0.008$1.28DeepSeek V4.1 Flash1.0M context$0.3$1.2DeepSeek V3.1 Terminus164K context$0.3$1DeepSeek V3.1164K context$0.25$0.95Batch, fast-mode and free variants are excluded. Prices change; the figure re-draws from the catalogue on every build.
Mistral AI API prices ($ per 1M tokens, live from our catalogue)Live from the OverpayingForAI catalogue · last verified 2026-10-02 · sorted by output priceInputOutputMistral Medium 3.5262K context$1.5$7.5Devstral 2 2512262K context$0.4$2Mistral Medium 3.1131K context$0.4$2Mistral Large 3 2512262K context$0.5$1.5Codestral 2508256K context$0.3$0.9Mistral Small 4262K context$0.15$0.6Mistral Small 3.2 24B256K context$0.094$0.25Ministral 3 8B 2512262K context$0.15$0.15Batch, fast-mode and free variants are excluded. Prices change; the figure re-draws from the catalogue on every build.
Meta models: hosted API price per 1M tokens (live from our catalogue)Live from the OverpayingForAI catalogue · last verified 2026-10-02 · sorted by output priceInputOutputMuse Spark 1.31.0M context$1.25$4.25Muse Glimmer 30B131K context$0.35$1.5Llama 4 Maverick1.0M context$0.188$0.652Llama 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.
xAI Grok API prices, $ per 1M tokensLive from the OverpayingForAI catalogue · last verified 2026-10-02 · sorted by output priceInputOutputGrok 4$3$15Grok 4.6500K context$2$6Grok 4.7500K context$2$6Grok 4.5500K context$2$6Grok 4.31M context$1.25$2.5Grok 4.202M context$1.25$2.5Grok Build 0.1256K context$1$2Batch, fast-mode and free variants are excluded. Prices change; the figure re-draws from the catalogue on every build.

AI agents

New autonomous agents, explained fast.

Short courses on the newest agent products — what they actually do, which plan or tier includes them, and how to write a task worth the run without wasting a subscription.

5 courses · 15 lessons

The agent loop — why agents cost more than chatsEach step is a full model call that re-reads everything so far. Ten tool calls can bill 10× a single answer.Goalone instructionModel plansreads full historytool callTool runssearch, code, filesresultModel reads resultanother billed callrepeat until done — cap the number of steps and the token budget
xAI·Beginner·3 lessons · 20 min

Grok Bot 101: What xAI's New Autonomous Agent Actually Costs

Grok Bot is xAI's answer to "an agent that works while you don't watch it." Three lessons on what it does, which SuperGrok tier it needs, and how to write it a task worth the run.

  1. 1What Grok Bot is, and what it isn't
  2. 2Which SuperGrok tier actually includes Grok Bot
  3. 3Writing a Grok Bot task worth the run
The agent loop — why agents cost more than chatsEach step is a full model call that re-reads everything so far. Ten tool calls can bill 10× a single answer.Goalone instructionModel plansreads full historytool callTool runssearch, code, filesresultModel reads resultanother billed callrepeat until done — cap the number of steps and the token budget
Vendor-neutral·Beginner·3 lessons · 19 min

OpenClaw 101: The Free Agent Gateway With a Hidden Token Bill

OpenClaw is a free, open-source bridge between Discord/Slack/Telegram and an LLM agent. Three lessons on what it is, what it actually costs once it's running, and setting up your first bot.

  1. 1What OpenClaw is, and why the name changed twice
  2. 2The real cost: whichever model you plug in
  3. 3Setting up your first OpenClaw bot
Open-weight models: three ways to run the same modelBecause the weights are downloadable, you are not locked to one vendor's price list.Official API
Vendor hosts it. Simplest, usually cheapest at low volume.
pay per token
Third-party host
OpenRouter, cloud providers. Competes on price and speed.
pay per token, often lower
Self-host
Your GPU or rented GPU. Fixed cost, no per-token bill.
hardware + ops time
Self-hosting only wins at sustained high volume. Below a few million tokens a day, a hosted API is almost always cheaper once you count your time.
Nous Research·Beginner·3 lessons · 20 min

Hermes 101: The Model, the Agent, and the Portal Aren't the Same Thing

"Hermes" from Nous Research names three different products with three different price stories. Three lessons on telling them apart, pricing the model, and running the free agent.

  1. 1Wait, which Hermes do you mean?
  2. 2Pricing the Hermes model: three ways to access the same weights
  3. 3Running Hermes Agent, the free self-improving coworker
What actually drives an AI billRoughly ordered by how often they surprise people.1Output length
Output tokens cost 3–5× input. Verbose answers and long reports dominate.
2Re-sent history
Every chat turn re-bills the whole conversation and any pasted files.
3Wrong tier
Flagship model on tasks a small model does identically.
4Retries
Vague prompts → three attempts → three bills.
5Idle subscriptions
Seats and plans nobody used this month.
6Agents & tools
Multi-step loops multiply calls silently.
GitHub·Beginner·3 lessons · 19 min

GitHub Copilot AI Credits 101: What 1 Credit Actually Buys You

Copilot's paid plans now include a monthly "AI Credits" allowance instead of an unlimited chat tier. Three lessons on what a credit costs, which plan gets you the best rate, and what still doesn't touch your balance.

  1. 1What an AI Credit is, and what still doesn't touch it
  2. 2Which plan gives you the best credit-per-dollar rate
  3. 3Making your credit allowance last the month
Subscription vs API: two ways to pay for the same modelFlat monthly fee with usage limits, or metered per-token billing with no ceiling.Subscription (chat app)API (pay per token)Fixed price every monthBill tracks real usageUsage caps and rate limitsNo cap — budget alerts are on youBest for daily heavy chatBest for bursts, automation, appsIncludes the app, files, memoryYou bring your own interfaceIdle months are wasted moneyIdle months cost $0Break-even test: if your API bill for the same work would be under the subscription price, the subscription is the overpayment.
Anthropic·Beginner·3 lessons · 19 min

Claude Code 101: Pro, Max and the API Bill Trap

Claude Code's plans are simple on paper — Pro, Max 5x, Max 20x. What trips people up is a single setting that quietly swaps your plan capacity for a separate, metered API bill.

  1. 1Claude Code's plans: Pro, Max 5x and Max 20x
  2. 2The trap: how a $20 subscription turns into an API bill
  3. 3When Max actually earns its higher price

Skills and decisions

Get better, spend less.

Cross-vendor courses on free tiers, the $20 question, reading the playground, agents without code, and spotting an overpay.

6 courses · 30 lessons

The free-first ladder1Free chat tierChatGPT, Claude, Gemini2Free API rows$0 per token, rate-limited3Cheap paid APIcents per month4$20 planonly when time is the cost
Vendor-neutral·Beginner·5 lessons · 31 min

Best Free AI in 2026: What the Free Tiers Really Include

Five short lessons on getting real work out of the free tiers of ChatGPT, Claude and Gemini, the free API models, and the exact moment $20 a month stops being a waste.

  1. 1What the free tiers of ChatGPT, Claude and Gemini actually include
  2. 2The free API models, and what they are actually good for
  3. 3Make a free tool do paid work
  4. 4The moment paying makes sense
  5. 5The free-or-paid decision checklist
Subscription vs API: two ways to pay for the same modelFlat monthly fee with usage limits, or metered per-token billing with no ceiling.Subscription (chat app)API (pay per token)Fixed price every monthBill tracks real usageUsage caps and rate limitsNo cap — budget alerts are on youBest for daily heavy chatBest for bursts, automation, appsIncludes the app, files, memoryYou bring your own interfaceIdle months are wasted moneyIdle months cost $0Break-even test: if your API bill for the same work would be under the subscription price, the subscription is the overpayment.
Vendor-neutral·Beginner·5 lessons · 32 min

Is ChatGPT Plus Worth It? The $20 AI Subscription Guide

The $20 tier is a category, not a product. Five lessons to pick the right one for your job, see what the same money buys on the API, and leave the moment it stops earning its keep.

  1. 1The $20 tier is a category, not a product
  2. 2Pick by job, not by brand
  3. 3What $20 buys, versus the API
  4. 4The cancel-and-switch checklist
  5. 5Make the call, and write it down
How the playground scores a model1One promptidentical text2Every modelin a family3Checksdid it follow the rules?4Judge0–10 for quality5Cost + latencyper run, from real usage
The same five moves, whatever your job1Paste your materialthe chapter, the ad, the…2Say the formatbullets, a table, 120 wo…3Run itfree tier first4Check one factthe model will not5Tightenshorter, in your voice
The agent loop — why agents cost more than chatsEach step is a full model call that re-reads everything so far. Ten tool calls can bill 10× a single answer.Goalone instructionModel plansreads full historytool callTool runssearch, code, filesresultModel reads resultanother billed callrepeat until done — cap the number of steps and the token budget
What actually drives an AI billRoughly ordered by how often they surprise people.1Output length
Output tokens cost 3–5× input. Verbose answers and long reports dominate.
2Re-sent history
Every chat turn re-bills the whole conversation and any pasted files.
3Wrong tier
Flagship model on tasks a small model does identically.
4Retries
Vague prompts → three attempts → three bills.
5Idle subscriptions
Seats and plans nobody used this month.
6Agents & tools
Multi-step loops multiply calls silently.

Questions before you start

Are these AI tutorials free?

Yes. Every 101 course on OverpayingForAI is free, has no sign-up, and works on any device. Progress is saved in your browser.

Which AI should a beginner learn first?

Start with the one you already have access to. ChatGPT, Claude and Gemini all have free tiers that cover the basics in these courses. The pricing lessons in each course show when paying is actually worth it.

Do the prices in the tutorials stay current?

The price figures are drawn from the same catalogue that powers the site's pricing tracker, so they redraw whenever the catalogue updates. Prose figures are marked as 'at the time of writing'.

How long does each course take?

About an hour of reading per course, split into eight lessons of six to twelve minutes. Each lesson stands alone, so you can jump straight to pricing, prompting or the API quickstart.

If our calculators helped you cut down on hidden AI wallet leaks, thanks for using them. A tiny fraction of your savings is what 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.