Chat Tools, Coding Assistants and App Builders

Why this matters

Beginners lose weeks to tool choice: signing up for five services, hitting a paywall mid-task, pasting code between windows that do not connect. The landscape looks crowded but it collapses into three categories, and this course needs exactly one thing from each. Learn the categories once and the noise drops away: every new product announcement becomes "which of the three is this, and do I need it", usually answered with "no".

Concepts

1. Chat tools. You type messages, you get text back. Input: words, plus optionally a file or image. Output: words, including code you copy yourself. Use them for: briefs, explanations, plans, error messages, drafts of any text. Examples you will meet in guides (verify current names and terms on their official pages before signing up): general chat assistants offered by the large AI providers, usually with a free tier and a paid tier above it.

2. Coding assistants inside an editor. They live where the files live. Input: your project files plus your instruction. Output: edits to those files, suggested directly where they belong. Use them for: changing real code once you have files (modules M04 to M07). They see the surrounding code, so their suggestions fit better than pasted chat answers, but they still need small, precise tasks. Many editors bundle an assistant; some are free within limits, some are paid. You do not need any of them for this course: every lab has a path that uses only the browser and the provided materials.

3. Browser-based app builders. You describe an app, they generate a working first version inside the browser: files, preview and hosting in one place. Input: a project description. Output: a runnable app you then refine by chatting. Use them for: the fastest possible first version of a small idea, and for seeing how description quality controls result quality. Limits are real: free tiers cap how much you can generate, and moving the result out to your own files takes deliberate export steps (tool guide 5 covers exactly that).

Plans, quotas and API keys, untangled. A plan is what you pay (or do not pay) for a product: free tier, monthly subscription, pay-as-you-go. A quota (also called credits or usage limit) is how much you may use: a number of messages, generations or tool calls per day or month. Quotas apply on free and paid plans alike; paying raises the ceiling, it rarely removes it. An API key is a secret string that lets your own program call a provider's model directly and bills you per use. You do not need an API key anywhere in this program, and you should never paste one into a chat, a prompt, or a file you share (lesson 4 shows the safe habits).

Worked example

Task: "I want a page where visitors can leave their email for bakery news." Which category? If you need words first (what should the page say, what should happen after signup), start with a chat tool and draft the brief. If you already have files and want the signup form added to them, that is an editing job for a coding assistant. If you want to see a working signup page in two minutes to react to, try an app builder, then export and continue in your own files. Same goal, three doors; the brief you write in M02 opens whichever door you choose.

Expected result: for any new task you can name the category in one sentence ("this is a briefing task, so chat tool") and state its cost honestly ("free tier, limited generations per day, checked on the official pricing page").

The common wrong move

Locking the whole course onto one specific product because a tutorial said so, then stalling when it demands payment or disappears. This program is tool-independent on purpose: concepts transfer, buttons do not. When a guide names a product, it also names the category and the free path, and every required lab works without any paid account.

A note on checking product facts

AI products change fast: names, prices and free limits move every few months. Treat any product detail in this course, including in tool guides, as a snapshot with a check date, and re-check the official documentation before you sign up or pay. The guides show you exactly which official page to open for each claim.

Lab and next step

No lab in this module asks you to create an account anywhere. Lab A03 trains the safety reflex instead: spotting what must never be shared. Next, lesson 4 collects the first safe working habits: passwords, keys, other people's data, and refusing to obey a model's risky instruction.