LuciQA / Features

Features

Screen capture, local OCR, image questions, and provider comparison in LuciQA.

Starts from the menu bar or system tray

LuciQA sits in the macOS menu bar or the Windows and Linux system tray. Choose OCR, Image or Multi there, then drag over the part of the screen you want to ask about: a chart, a code snippet, a formula or a study question.

Answers can be wrong. The confidence value is the model’s own estimate, not a measured probability. Check anything that matters.

Use LuciQA only where AI assistance is allowed, and follow the rules of your school, employer, the website and the provider. AI answers can be wrong.

Read the guide →

Answer window: one provider answers a sample physics question with a worked formula and a confidence estimate. — Sample content

Text when it is text. Image when it is visual.

OCR runs on your device using bundled recognition data for all nine supported app languages. For formulas, charts and diagrams, Image mode preserves the selected visual region. Either way, review what you send.

For text questions, OCR runs locally. The selected text is then sent to your configured provider. Image mode sends the selected image.

Answer window in Image mode: a sample bar chart was sent and the answer names the two months with the largest increase. — Sample content

Your providers. Your keys.

Connect the providers you already use. LuciQA sends selected content directly to them and does not resell AI credits or relay requests through its own server.

Groq · Gemini · OpenRouter · Cloudflare Workers AI · Cerebras · Mistral · OpenAI · Claude

Your keys are saved in local app settings. Secure OS credential storage is required to use them.

View all providers →

Free provider quotas

Groq, Gemini, OpenRouter, Cloudflare Workers AI and Mistral currently offer free API quotas, so LuciQA Free can be used at no cost within those limits. Each provider sets and may change its limits. Cerebras offers a time-limited trial; OpenAI and Claude are paid APIs.

Check current limits in your provider account.

Free AI APIs for questions on your screen →

Answer style, thinking level and theme

Choose answer only, or a short, balanced or detailed explanation. Set the thinking level separately for each model that supports it. Light and dark themes are included.

Nine languages

The interface is available in English, German, French, Spanish, Brazilian Portuguese, Turkish, Russian, Japanese and Korean. On-device OCR reads the same nine languages.

Settings in the dark theme: four explanation lengths, the “If needed” confidence threshold and the custom system instruction field. — Sample content

Pro

Compare answers

One provider answers in Free. Pro can ask several providers and call additional ones when answers need checking.

Every question

Set several providers to “Every question”. They are asked at the same time and each answer is shown on its own, so you can see where they agree and where they differ.

Answer window with two providers giving the same result for a sample code snippet, marked “2 models agree”. — Sample content

“If needed” providers

A provider set to “If needed” is asked only when the other answers differ, a request fails or returns nothing, or the reported confidence is below your threshold.

Answer window where two providers gave different answers and a third, set to “If needed”, was asked as well. — Sample content

Multi

Pro Multi suggests up to 12 question boxes on one page. Move, resize, add or remove boxes, choose OCR or Image for each, and edit the OCR text before anything is sent. Each question is a separate request; two run at a time.

Reference document

Import a .docx reference. Its extracted text accompanies each question sent to your providers and uses their token allowance.

Custom system instruction

Add an optional system instruction for every queried provider. The answer layout stays the same.

Reference document settings showing a sample .docx file and its estimated token use per question. — Sample content

A clear data path

API keys live in local app settings with OS secure storage. Selected text, images and any Pro document or instruction go directly to your configured AI providers. Their policies and costs apply.

macOS · Windows · Linux

Find LuciQA installers and system requirements for macOS, Windows, and Linux.

Pro · One-time

Pro is a one-time purchase: a perpetual license for LuciQA 1.x on up to three devices. There is no subscription, and AI usage is not included.

Pricing →