AI tool tier lists by category

AI that writes, fixes and reviews code. The choice comes down to autocomplete vs. agent, which editor it plugs into, and whether there is a plan an individual can afford.

General-purpose assistants you talk to — questions, drafts, reading documents. They mostly do the same job; fresh-data access, free limits and how much context they hold are what separate them.

Images from a prompt. Style consistency, how it renders text, and free generation limits are where these pull apart.

Tools that find sources and organise evidence. The fork: whether it actually cites, and whether it reads only your uploads or searches the web too.

AI for notes, meetings, scheduling and documents. Whether it plugs into what you already use — calendar, Notion, Slack — decides if it sticks.

AI for slides, logos and UI mockups. Some generate from scratch, others let you edit real files — decide how much you need to tweak the output before you pick.

Drafting help for fiction and scripts. It comes down to whether it remembers a long story's setup, and where its content limits sit.

Video from a prompt or a single still. Clip length, watermarks and how fast credits burn decide how much you can actually try for free.

Help with paper drafts, citations and literature reviews. Above all: whether it fabricates citations, and which reference styles it supports.

AI for sentences, documents and whole pages. Quality swings hard by language pair, so weigh the free character cap and whether formatting survives.

Engines that answer instead of listing links. Trust comes down to whether sources are cited and how fast the index catches up with what is new.

Tools for writing content built to rank. Draft speed matters less than whether real keyword data is built in — and how much editing the output still needs.

Cutting long documents, videos and meetings down. What matters: how much you can feed it at once, and whether it invents things the source never said.

Text to speech, voice cloning and transcription. What decides it in practice: how natural your language sounds, and what the licence allows commercially.

Describe an app, get one. Check first whether you can export the code, and whether hosting and a database come attached.

Songs from a genre and some lyrics. Commercial rights and whether you get separated stems is what splits hobby use from real work.

Ad, landing page and email copy. The fork: whether it gives you several angles at once, and whether you can teach it your brand voice.

Rewriting a sentence without changing what it means. Look at whether tone control actually works, and the free character cap.

Spelling and grammar checks. Whether it handles your language properly and lives in your browser or Word decides how often you actually use it.

Tools for fixing images you already have — background removal, upscaling, object edits. In practice the gap is less about quality than free-tier limits and watermarks.