FEATURES

Everything that makes Discria special.

Adaptive quiz engine, your own sources, local LLM — all of Discria's capabilities at a glance.

ADAPTIVE ENGINE

Learns the way you learn.

Discria uses statistical, individual, and AI signals to measure which questions actually hold you back — not just how often you got them wrong.

Then it asks: is it you — or the wording of the question? Discria creates two variants of the same question (A/B), presents both, and permanently keeps the one where you do better. The weaker one is discarded.

No classic flashcard deck can do this — it just shows wrong cards more often. Discria adapts the questions themselves to your learning style.

TOPICS & SOURCES

Your material. Your quizzes.

You create topics — from operating systems to Spanish vocabulary — and attach your own sources per topic: folders, individual files, or hefty textbook PDFs.

Optionally, Discria can independently perform web research to enrich your topic with current knowledge. From this it generates multiple-choice questions — based entirely on your material, not on a third-party content library.

Folders & files

Your own scripts, notes, lecture summaries — whatever you have.

Textbook PDFs

Even large, multi-hundred-page PDFs are fully processed.

Web research (optional)

Discria can search for current context on your behalf.

LEARNING GAME

Learning that feels like a game.

Discria generates real multiple-choice questions with single- and multiple-answer formats — automatically from your sources, immediately playable as a learning game.

A progress bar and a running streak keep you on track. Discria deliberately selects the questions where you still need work — and rephrases them when needed.

LOCAL & PRIVATE

Your material stays yours.

Discria runs preferably with a small local model via Ollama — entirely on your machine, no cloud required. Your materials never leave your computer.

If you prefer your own API endpoint, you can optionally connect LiteLLM. In either case: no account, no telemetry, no dependency on external services.

A local device with a shield and lock — your data stays inside and never leaves your machine.

Ollama (local, preferred)

Small local model runs directly on your device.

LiteLLM (optional)

Your own API endpoint — you stay in control.

No account, no cloud

No mandatory sign-up, no telemetry.

PRE-RELEASE

Discria is in development.

All the features above are on the roadmap for early access. Claim your spot now — you'll be the first to know when Discria launches.