A writing tool with a built-in editor — not the kind that writes for you, but the kind that reads what you wrote.
Sign up or log in — it takes thirty seconds.
You're looking at a Lucid document right now. The toolbar above, the title, the clean layout — it's the same editor your drafts will live in. Go ahead, type something. Nothing here is saved.
You write. An LLM reads back.
When you're ready for feedback, you hit Critique. The LLM reviews your draft and leaves margin comments — each one anchored to a specific sentence, like a careful editor's pencil marks. You choose how nitpicky it should be, from "only flag what's broken" to "pick every nit."
You work through the comments one by one: revise the sentence, reply to push back, or archive the thread and move on. If the LLM misreads your intent, tell it — it replies in the same thread, right next to the text you're both looking at.
The AI never silently rewrites a sentence. It can only point at your words and make a case for changing them. You decide what to do about it.
Most AI writing tools are draft factories. Type a prompt, get prose. The author's job becomes accepting or rejecting someone else's writing.
Lucid inverts the relationship. You are the author. The LLM is a reviewer — opinionated, sometimes wrong, always forced to show its work. Every suggestion lives in a thread you can engage with or ignore. Your voice stays yours. The result is better than what you'd have produced alone.
Do I need to pay for an API key? Lucid uses your own Anthropic API key, so you're in control of your usage and costs. You'll add it after signing up. A typical critique run costs a few cents.
What kind of writing is this for? Anything long-form: blog posts, essays, articles, documentation, fiction. If you'd benefit from a second pair of eyes before publishing, Lucid is useful.
What happens to my documents? Your documents are private. Only you can see them.