FAQ
Common questions
Everything you want to know about how Soro Soke works. If your question isn't here, get in touch.
The first 50 accounts are completely free — forever. No card, no trial, no expiry. If there are still founder spots available when you sign up, your account is yours for life. Once those 50 spots fill, new accounts get a 14-day free trial with full access before choosing Pro at $9.99/month or $99/year.
Spaced repetition is a learning technique that schedules reviews at increasing intervals over time. Instead of reviewing everything every day, you only review a card when you're about to forget it. Soro Soke uses the SM-2 algorithm — the same one that powers Anki and SuperMemo — to calculate these intervals automatically.
When you type a word or concept into a new card, Soro Soke generates a definition, a memory hook, an example sentence, and synonyms in seconds. You can edit any of it before saving. The goal is to cut the time it takes to create a good card from minutes to seconds.
Free recall means retrieving information from memory without any cues — like writing down everything you remember from a deck before you flip any cards. Research consistently shows it produces stronger memory than passive re-reading. Soro Soke makes it a first-class part of your review session.
Yes. Soro Soke works for any vocabulary in any language. Soro Soke's card drafting supports multiple languages — just type the word in the language you're studying and it will respond accordingly.
You can add cards one at a time (with Soro Soke's drafting), or import a CSV file in bulk. Each deck has an export button too, so you can back up your cards or move them between accounts.
Deleting your account permanently removes all your decks, cards, review history, and profile data. You can do this from Settings → Delete account. It requires typing your email address to confirm — there is no undo.
Soro Soke is a web app that works in any modern browser — Chrome, Safari, Firefox, Edge. It's fully mobile-responsive, so it works well on your phone or tablet without needing to install anything. A native iOS and Android app is on the roadmap.
Partially. If you lose your connection mid-review, Soro Soke detects it and holds the grading buttons until you reconnect — so you never accidentally lose a session. Features that require AI (card drafting, Speak Up, Small Talk Lab) need an active connection. Full offline review caching for commutes is on the roadmap.
Yes. Your dashboard shows your current review streak, a 30-day activity heatmap, total cards mastered, and the next 8 days of upcoming reviews so you can see what's coming. Consistency over long periods matters more than any single session — the streak is there to reflect that, not to guilt you.
That depends on how many cards you have and how well you know them. Soro Soke introduces 3 new cards per day by default. Most consistent users have a daily queue of 10–25 cards, which takes 5–10 minutes. The key is not letting the queue pile up — daily short sessions beat infrequent long ones by a significant margin.
Nothing catastrophic. Your due cards wait for you. If you miss several days, more cards will be due when you return — Soro Soke prioritises overdue cards first so you work through the backlog efficiently. Your streak resets, but your card mastery and review history are untouched.
There are six grades, each telling the algorithm how strongly you recalled the card. Blackout (0) — no memory at all, card returns within minutes. Wrong (1) — incorrect recall, returns shortly. Hard (2) — correct but required serious effort, interval shortens. Okay (3) — correct with hesitation, kept close to its current schedule. Good (4) — correct with only a brief hesitation, interval extends. Perfect (5) — instant, effortless recall, interval extends significantly. Grading yourself honestly — not generously — is what makes the algorithm work well over time.
New means the card hasn't been reviewed yet. Learning means you've reviewed it but it's still in early repetition — coming back frequently. Familiar means consistent good recall over several sessions — intervals are stretching into weeks. Mastered means strong, reliable recall across many sessions — the card surfaces infrequently, maybe monthly. Cards move up and down based on your grades.
Yes — deck sharing is live. From any deck page, generate a share link. Anyone with the link can preview the cards and clone the entire deck to their own account in one click. You can revoke the link at any time.
Your sessions are private. Speak Up conversations and Small Talk Lab exchanges are not stored beyond the session — they exist to give you feedback, not to build a record. Your card data is tied to your account and never shared or sold.
Anki uses the same SM-2 algorithm as Soro Soke and is powerful — but it requires you to build everything yourself and has a steep learning curve. Quizlet is primarily a recognition tool with limited scheduling depth. Soro Soke adds Soro Soke-drafted cards (no blank-card friction), Speak Up for high-stakes speaking practice, and Small Talk Lab for social fluency. It's built for professionals who want to communicate better — not students drilling for exams.
Speak Up is for high-stakes prepared scenarios — a job interview, a salary negotiation, a pitch, pushing back on a decision. You rehearse before a specific kind of moment happens. Small Talk Lab is for social fluency — you open a conversation in a natural setting (a networking event, a long flight, a dinner party) with no script or guided prompts, then get coaching on your conversational moves. One builds professional confidence under pressure; the other builds social ease.
After a Speak Up session you receive a score from 0–10, a breakdown of what landed well, specific areas to sharpen, a filler-word count tracking verbal habits like 'um' and 'uh', a model response showing how your answer could have landed better, and an idealized version of the full conversation — how it could have gone if you had been at your best from the start. The score reflects clarity, structure, how well you held your position, and your verbal delivery.
Debate Lab puts you in a structured debate against an AI opponent. Pick a formal motion (like 'AI should be regulated by governments') or describe a real-life situation — a salary push, a product decision, a disagreement with a colleague. Choose your position, study the prep room, then go five exchanges with an AI that argues back and adapts to your points. Afterward you receive a score, argument-by-argument feedback, skill sub-scores across five dimensions, and a full coaching debrief.
Speak Up is one-sided and coaching-focused — you rehearse delivering a response (a pitch, an interview answer, a negotiation opener) and the AI evaluates it. Debate Lab is adversarial and interactive — the AI argues back, challenges your logic, and tries to win. It's the difference between rehearsing a speech and actually being challenged in a meeting.
After five exchanges you receive an overall score from 0–100, a Win/Draw/Loss verdict, and skill sub-scores across five dimensions: Clarity, Evidence, Rebuttal, Logic, and Composure. You also get an argument-by-argument breakdown (each exchange rated Strong, Okay, or Weak), your opponent's strongest point, and the best counterargument you missed.
Before the debate begins, the prep room gives you three key arguments to lead with, two likely attacks your opponent will make (each with a suggested rebuttal), and one thing to watch out for based on your motion and the opponent type you chose. Read it, prepare, and enter when you're ready.
The sway meter tracks how the imagined audience's opinion shifts in real time across each exchange. A strong point swings the meter toward you; a weak or conceded point swings it back. By the end you can see the cumulative arc of the debate and which exchanges moved the needle most — not just your final score.