# Kunkafa - full text > Generated at build from the same sources as the pages. Canonical pages: https://kunkafa.com ## Home Kunkafa – Every market's expected move, with the odds Gold · 12 days Example Could Gold go up 1% within 12 days? 84% chance of reaching it, say Kunkafa's models ↓ Down 0.6% in the same period: 70% chance Either, both or neither could happen. Kunkafa simplifies Market Reading Tired of trying to make sense of the market with dozens of indicators? Feeling overwhelmed? Thinking gold is going up forever? Afraid of a market collapse? Kunkafa helps you think through every scenario: probabilities for what comes next, not just a chart of what came before. Start your 7-day free trial See how it reads Probability, Not Prophecy. Forecast, not advice. 5 min → 7 years Forecasts cover your desired horizon 70 AI experts Diverse views of the same market Both directions Assess the scenarios and probabilities Two voices. One map. Nasdaq going up forever? That's greed talking. Fear says the opposite. Both can wreck the next decision. Kunkafa doesn't take a side. It maps the paths, how far each one goes, and the chance of each. Both directions. From five minutes to seven years. With the record beside it. See the full picture. Assess the risk, not just the reward. Forecast, not advice. Built for rational thinkers who take emotion out and think in probabilities and scenarios. Greed: up forever Fear: a collapse +0.1% within 1 hour: 78% chance +0.3% within 1 day: 81% chance +1% within 12 days: 84% chance +2.5% within 50 days: 79% chance +5% within 6 months: 74% chance +9% within 2 years: 68% chance −0.1% within 1 hour: 75% chance −0.2% within 1 day: 77% chance −0.6% within 12 days: 70% chance −2% within 50 days: 73% chance −4% within 6 months: 71% chance −7% within 2 years: 66% chance NOW 5 minutes 2 years up down Greed draws one line up. Fear draws one line down. Kunkafa draws every path, with the chance of each. Dots: the gold example, one hour to two years. The product One forecast. Three ways to read it. From one question to the full view. Same numbers, three readers. Level 1 The question Your market, your duration, a round threshold; the chance, then the other side below it. Level 2 The picture Every move from one point: right is more time, up or down is a bigger move, solid dots are likelier. Level 3 The full view The price line, the forecast band, and every level with its chance. Gold Example 01 · Market Gold 02 · Duration 12 days 03 · Distance 1% Could Gold go up 1% within 12 days? 84% chance of reaching it ↓ Down 0.6% in the same period: 70% chance Either, both or neither could happen. The other side sits below the answer, one glance away. Try it yourself: 7-day free trial Example figures for Gold, chosen to be easy to read. The app shows its own. Forecast, not advice. Results Measured, not promised Trained on about 10 billion data points. Tested on about 2 billion it had never seen. Then counted, every day, on live markets: how past forecasts turned out, at the strength the app uses by default (75%). Updated 7 September 2026. Finished in the forecast direction Last day 85% 140,652 evaluated Last 7 days 84% 8,793,092 evaluated Last 30 days 84% 9,876,426 evaluated All time 84% 9,876,426 evaluated By forecast strength finished in the forecast direction · All time 50% 75% 100% 5% 25% 50% 75% 95% 84% at 75% Read how it works → · See the results page → Past results do not promise future ones. Up and down both remain possible. The Problem With AI In Finance Why LLMs Fail at Financial Prediction ChatGPT, Claude, and Gemini are extraordinary at language. But asking them for stock predictions is like asking a poet to do calculus. LLMs (ChatGPT, Claude, Gemini) Designed for text, not numbers Language models process words, not price patterns in tabular data Hallucinate certainty They give confident-sounding answers even when they have no basis for prediction Trained on articles about markets Not on actual price data - they've read opinions, not charts Cannot quantify uncertainty No native probabilistic output - just text that sounds confident "The stock will go up because of positive momentum..." - A typical LLM response (meaningless) Kunkafa Built for numbers, not words Rows and columns of price history, not articles about it Says how likely, not just which way Every level carries its probability, and both directions are shown Trained on real price history About 10 billion data points across stocks, indices, currencies, commodities and crypto Seventy experts, not one When the experts disagree, the probability drops, and you see that too "Price rises 0.9%: 61% probability. Price falls 1.6%: 55% probability." - A Kunkafa forecast: both directions, each with its probability At A Glance LLMs Kunkafa Designed For Text & language Numbers Training Data Articles about markets Actual market data Output Style "The stock will go up" "Price rises 0.9%: 61% probability. Price falls 1.6%: 55% probability." Uncertainty Hallucinates certainty Gives each outcome a probability Self-Awareness Doesn't know what it doesn't know Says how sure it is It's about the right tool for the job . LLMs are extraordinary for language. Kunkafa is purpose-built for tabular prediction. The Kunkafa Approach Purpose-Built for Uncertain Domains Markets are noisy and unpredictable. We built AI that embraces it. Built for numbers Our models learn from the rows and columns of price history that define markets. Every forecast says how sure it is A 52% is shown as a 52%, and a near-even split is the honest answer, not a failure to find one. The Proving Ground Markets are the ultimate test — noisy, adversarial, and unforgiving. If it works here, it works anywhere. Probability, Not Prophecy Plans Start free. Looking at a forecast costs nothing and moves no counter. All paid plans include a 7-day free trial For the curious Free Free Every market in the catalogue, on the shortest forecast Follow up to 5 markets One alert, so you can stop watching the screen The full record of how past forecasts turned out Start free For the focused Starter $10 / month 7-day free trial One market of your choosing, on every forecast duration Every other market stays on the shortest forecast Follow up to 10 markets Up to 5 alerts Start the 7-day trial For the ambitious Core $49 / month 7-day free trial Every market, on every forecast duration Follow up to 25 markets Up to 20 alerts Wider evidence, and a home screen shaped around what you follow Start the 7-day trial For the committed Pro $99 / month 7-day free trial Everything in Core Follow up to 100 markets Alerts without a cap Automated trading, with limits you set and can pause Start the 7-day trial Free is where every account starts. No trial that ends. Secure payments Cancel any time Powered by Stripe --- ## How it works How Kunkafa works out the chance a market reaches a price How it works How Kunkafa works out the chance a market reaches a price One question, asked of every market: how far could the price move, in which direction, and how likely is each outcome, across every duration at once. This page explains where the answer comes from and the rules it follows. By Kareem Farid , founder. Updated 7 September 2026. Seventy experts, one question each Seventy independent views of the same market. When they agree you see it; when they disagree, that tells you something too. Each expert looks at the same market and answers on its own: how far, and which way. There is no single model with a single opinion. The chance you see for a level is not one voice; it is what the record shows for forecasts like this one, and the experts’ agreement or disagreement is inside it. A level that many experts reach the same way carries a higher chance; a level they split on carries a lower one, and you see that. We do not publish which methods the experts use or how they were built. What we publish is how their forecasts turned out, every day. What a forecast is made of Every forecast is four things, and always four: how far (the change from today’s price, as a percentage), which way (up or down), how long it has (the duration), and the chance it gets there . The side you asked about comes first; the other side sits below it. A forecast that hid one side would be claiming the price cannot go there, which is false. On the screen the four things are a ladder: one row per level, with its change, its duration and its chance, above a price line that draws the same levels as a band. Where the chance comes from The chance beside a level is a rate from the record: of past forecasts like this one, at this strength, how many reached the level. It is measured against what actually happened, not estimated from a formula about volatility. That is the difference between Kunkafa and an expected-move calculator: the calculator tells you a range the options market implies; Kunkafa tells you how often the price really got there. The experts’ stated chances are checked against outcomes and corrected, so that a level shown at 70% is one that forecasts like it reached about seven times in ten. In the record you can see this for yourself, by strength. Duration is a ceiling, not a choice The slider under the market runs from five minutes to seven years, and its handle is a ceiling. “Up to 7 years” shows every duration at once, pooled into one picture; slide it down and the longer durations drop out. The bands under the track, Scalping, Day trading, Swing, Investing, sit where those durations fall, so the scale explains itself. Everything downstream pools the same way: the ladder, the price band, the record and sharing all cover every duration up to the ceiling. The record and its rules Beside every forecast sits “How similar forecasts performed”: how many forecasts were evaluated, how many were strong enough for your filter, how many finished in the forecast direction, and how many reached the level. It is counted over the last day, 7 days, 30 days, or all time, and it follows four rules: Under 100 finished examples, no rate. The record says “Not enough past examples to estimate performance reliably.” instead of showing a number nobody should trust. One filter drives everything. The strength you set narrows the list, and the record is counted on the same setting, so a number on screen never disagrees with the control above it. Filters narrow the list, never the drawing. The chart always shows every level. A rate under 1% says “under 1%”. One percentage format, and every percentage says what it measures. Each past forecast ends one of five ways: it reached the level, it passed it, the period ended ahead, the period ended behind, or it is still running. The results page publishes the same counts for every market, refreshed each morning. Training and testing, in scale The experts were trained on about 10 billion data points of price history across stocks, indices, currencies, commodities and crypto, and each was checked on millions of past moves it had never seen. After training, their stated chances were compared with what happened and corrected, so that the chance you read means what it says. That is all we publish about the internals, on purpose. What we publish instead, and what you should hold us to, is the record. What we never do No advice. A forecast is a chance, not an instruction. Forecast, not advice. For the rational investor: emotion out, scenarios in. Nothing simulated. An empty market says it is empty; a missing value is left blank and named, never filled with a plausible number. Nothing metered. Looking at a forecast costs nothing and moves no counter. No red and green, no countdowns, no streaks, no celebration. Up is mint, down is fuchsia, and nothing rewards watching the screen. No hiding one side. Up and down: the side you asked about first, the other always within reach. Glossary Expected move How far a market’s price could move, up or down, over a duration. Kunkafa shows several levels of expected move per market and duration, each with the chance of reaching it. Chance of reaching How often past forecasts like this one actually reached the level, measured against what happened. It is a rate from the record, not an opinion. Duration ceiling The duration slider’s handle. “Up to 7 years” shows every duration from five minutes to seven years at once; lowering it removes the longer ones. Forecast strength A filter on how strong a forecast has to be to appear in the list and count in the record. It narrows the list, never the drawing. Seventy experts The seventy independent judgements behind each market’s forecast, each giving how far and which way on its own. Agreement and disagreement both show in the chance. Finished in the forecast direction A past forecast that ended, at the close of its own duration, on the side it forecast. Reached the level A past forecast whose price touched the forecast level at some point within its duration, whether or not it stayed there. Still running A past forecast whose duration has not ended yet, so it has no outcome. Open Kunkafa Predictions See today’s results --- ## Results Kunkafa results, updated daily: how the forecasts turned out Results How Kunkafa’s forecasts turned out, updated daily Every morning we count how past forecasts ended, across every market and every duration the pipeline publishes, at the strength the app uses by default (75%). Updated 7 September 2026. Same definitions as the app’s record. Past results do not promise future ones. Up and down both remain possible. By window Window Forecasts evaluated Strong enough Finished in the forecast direction Reached the level Last day 140,652 46,647 · 33% 85% 84% Last 7 days 8,793,092 3,158,980 · 36% 84% 83% Last 30 days 9,876,426 3,261,235 · 33% 84% 83% All time 9,876,426 3,261,235 · 33% 84% 83% By forecast strength The stricter the filter, the fewer forecasts it counts and, usually, the higher the rate. How selective you are matters more than the rate itself. Strength at least Forecasts Finished in the forecast direction Reached the level 5% 9,314,058 79% 75% 10% 8,954,015 79% 76% 15% 8,648,635 80% 77% 20% 8,309,575 80% 77% 25% 7,955,240 81% 78% 30% 7,644,396 81% 79% 35% 7,273,395 82% 80% 40% 6,872,236 82% 80% 45% 6,441,560 83% 81% 50% 5,952,252 83% 82% 55% 5,449,321 83% 82% 60% 4,939,704 84% 82% 65% 4,404,924 84% 83% 70% 3,852,216 84% 82% 75% 3,261,235 84% 83% 80% 2,650,676 84% 83% 85% 1,966,834 84% 83% 90% 1,166,386 84% 84% 95% 251,760 86% 86% How to read this Forecasts evaluated : every forecast made in the window, whether it was strong or not. Strong enough : the ones at or above the strength shown, and their share of all forecasts. Finished in the forecast direction : of those, how many ended on the side they forecast at the close of their own duration. Reached the level : of those, how many touched the forecast level at some point within their duration. Under 100 finished examples a window shows no rate: “Not enough past examples to estimate performance reliably.” A rate under 1% reads “under 1%”. Definitions in full, and where the chance comes from, on the how it works page. The raw figures are published at /api/stats.json . --- ## FAQ Kunkafa Predictions FAQ – what the numbers mean FAQ Questions, answered straight What the app shows, what the numbers mean, and what we do not claim. 01 What Kunkafa is What is Kunkafa? Kunkafa is an AI research company working on prediction from numbers rather than words. Kunkafa Predictions is its first product: one screen that shows, for a market you pick, how far the price could move, in which direction, and the chance of each, across every duration at once. Is this financial advice? No, ever. A forecast is a chance, not an instruction. Nothing here tells you what to buy, sell or hold, and no result changes that. Forecast, not advice. For the rational investor: emotion out, scenarios in. Why “Probability, Not Prophecy”? Because nobody knows the future, and we would rather show you the odds of each outcome than pretend to. Most forecasts sit near 50/50, and the app says so. 02 What you see What does a forecast look like? A list of levels the price could reach, up and down, each with three things: the change from today's price, the duration it has, and the chance it gets there. Above it, the price line and the same levels drawn as a band. What does “up to 7 years” mean? The duration slider is a ceiling, not a choice. “Up to 7 years” shows every duration from five minutes to seven years at once; slide it down and the longer ones drop out. Why do I see up and down together? Because both are always possible. A forecast that hid one side would be claiming the price cannot go there, which is false. Why is it so often close to 50/50? Markets are efficient: whatever is known is already in the price. A near-even split is the honest normal answer; the app brings you the few forecasts that are not. What is forecast strength? A filter. “Only show forecasts at least this strong” hides the weaker ones from the list, and the record below is counted on the same setting, so a number on screen never disagrees with the control above it. It narrows the list, never the drawing. What are “How far it has to go” and “Both / Up / Down”? Two more filters: the smallest move you care about, and the direction. Like strength, they narrow the list and leave the picture alone. 03 Reading the record How do I know if the forecasts are any good? The record sits beside every forecast: “How similar forecasts performed”, with how many were evaluated, how many were strong enough for your filter, how many finished in the forecast direction, and how many reached the level. It is counted over the last day, 7 days, 30 days, or all time. What happens with only a few examples? Under 100 finished examples the app says “Not enough past examples to estimate performance reliably.” instead of showing a rate. A rate under 1% reads “under 1%”. What counts as a forecast turning out? Each past forecast ends one of five ways: it reached the movement level, it passed it, the period ended ahead, the period ended behind, or it is still running. How fresh is the data? Every forecast shows “Updated N minutes ago”. If a market stops publishing, the app says “No new data for N days” rather than showing old figures as current. 04 How it works How is the chance calculated? Seventy independent views of the same market. When they agree you see it; when they disagree, that tells you something too. Each expert judges how far and which way on its own; the chance you see is how often past forecasts like this one actually got there, checked against what happened. How were the experts trained? On about 10 billion data points of price history across stocks, indices, currencies, commodities and crypto, and checked on millions of past moves they had never seen. We do not publish the internals. Which markets are covered? Every market the pipeline publishes, listed in the app's Markets tab and on the results page . New markets appear as soon as they are published. 05 Plans and billing What is free? Every market on the shortest forecast, following up to five markets, one alert, and the full record of past forecasts. No trial that ends. What do Starter, Core and Pro add? Starter: one market of your choosing on every duration, follow ten, five alerts. Core: every market on every duration, follow 25, twenty alerts. Pro: everything in Core, follow 100, alerts without a cap, automated trading with limits you set and can pause. Paid plans start with a 7-day free trial. Compare the plans . How do I cancel? From Billing in the app, through Stripe, at any time. 06 Alerts, saving, sharing, automation What is an alert? One sentence you set: “Tell me when this market's estimated chance of moving up or down over a duration reaches N%.” It is checked on every bar close, on our side, so you can stop watching the screen. Can I keep or share a forecast? Yes. “Save this outlook” keeps a frozen copy in Activity; sharing publishes a card of the chart with a link back to the app, up to five a day. What is automated trading? On Pro, a rule that places a trade on your own exchange account (Binance, Bybit or OKX) when a forecast meets the condition you set, with the size, leverage and limits you set, and a pause switch. You complete a risk walkthrough before it is available. 07 Privacy and account What do you collect? Your account email, your plan, what you follow and save, and, if you accept the notice, how the site and app perform. Nothing is used for advertising. See the Privacy Policy . Can I delete my account? Yes, from Settings in the app, under Danger Zone; it removes your data. How do I reach you? Write to contact@kunkafa.com , or use the contact page . Disclaimer: Kunkafa provides probability estimates for informational purposes only. Nothing on this page constitutes financial advice. Past accuracy on validation data does not guarantee future performance. Both outcomes always remain possible. Methodology Performance Blog Try Kunkafa Predictions --- ## Contact Contact Kunkafa Contact Contact Ask about a forecast, a plan, or your account. We answer within two working days. Email The fastest way to reach a person. contact@kunkafa.com Community Other readers, and us, in the open. Join the Discord community Follow @Kunkafa_app on X What to include The market, the duration you were looking at, and a screenshot if you have one. For account questions, the email you signed up with. We can explain a forecast. We do not tell you what to do with it. --- ## About the author Kareem Farid, founder of Kunkafa About Kareem Farid, founder of Kunkafa Kareem Farid is the founder of Kunkafa. Before Kunkafa he founded Kundera (2019–2022), the first Arabic large-language-model app, built so researchers, media and students could put AI to work years before the big labs made it ordinary. He writes here about probability, markets and honest AI. What Kunkafa is Kunkafa is an AI research company working on prediction from numbers rather than words. Kunkafa Predictions is its first product: one screen that shows, for a market you pick, how far the price could move, in which direction, and the chance of each, across every duration at once. Seventy independent views of the same market; when they agree you see it, and when they disagree, that tells you something too. What it is not Not a trading terminal. Not signals. Not advice. Forecast, not advice. For the rational investor: emotion out, scenarios in. Open Kunkafa Predictions · Read how it works · Follow @Kunkafa_app on X --- ## Terms Terms of Service – Kunkafa Terms Terms of Service What using kunkafa.com and Kunkafa Predictions means, in plain words first. Last updated 7 September 2026. 1. Who we are Kunkafa Limited (“Kunkafa”, “we”) operates kunkafa.com and the Kunkafa Predictions app at predictions.kunkafa.com (the “Service”). Questions about these terms go to contact@kunkafa.com . 2. What the Service is Kunkafa Predictions shows, for markets it publishes, how far the price could move up or down over a range of durations, and the chance of each move measured against what happened in the past. It also offers alerts, a record of past forecasts, shareable forecast cards, and, on the Pro plan, automated trading rules that run on your own exchange account. By using the Service you accept these terms. If you do not accept them, do not use the Service. 3. Forecasts are chances, not advice A forecast is a chance, not an instruction. Nothing in the Service is financial, investment, legal or tax advice, and nothing recommends that you buy, sell or hold anything. Both directions remain possible in every forecast. Decisions you make, and their results, are yours. Past results, including the record we publish, do not promise future results. Markets can move against any forecast at any time. 4. Your account One account per person, with an email address you control. You are responsible for keeping your sign-in details safe and for what happens under your account. You can delete your account at any time from Settings in the app; deletion removes your data as described in the Privacy Policy . 5. Plans, trials and payment The Service has four plans: Free, Starter, Core and Pro. Paid plans are billed monthly through Stripe and start with a 7-day free trial; you can cancel at any time from Billing, and a cancellation takes effect at the end of the current period. Prices and what each plan includes are shown on the plans page and may change with notice; a change never applies before your next billing period. We do not store your card details; Stripe does. Refund requests are handled case by case: write to contact@kunkafa.com . 6. Automated trading (Pro) An automation is a rule you write: when a forecast meets the condition you set, an order is placed on an exchange account you connect, with the size, leverage and limits you set. You can pause every rule at any time. We execute the rule you wrote and nothing else. You are responsible for the exchange account, its keys, its funds, and every order the rule places, including losses. Automation is available only after you complete the risk walkthrough in the app. 7. Acceptable use Do not scrape, copy or resell the forecasts or the Service; do not attempt to bypass plan limits or security; do not use the Service for anything unlawful. We may suspend or end an account that breaks these terms. 8. Intellectual property The Service, its content, design and models belong to Kunkafa or its licensors. You may use the Service for your own purposes and share cards the app creates for sharing; anything else needs our written consent. 9. Liability In plain words: we provide the Service as it is, we do not promise it will be accurate, available or suited to any purpose, and we are not liable for losses from using it. THE SERVICE IS PROVIDED “AS IS” AND “AS AVAILABLE”, WITHOUT WARRANTIES OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. TO THE FULLEST EXTENT PERMITTED BY LAW, KUNKAFA, ITS DIRECTORS, EMPLOYEES AND AGENTS ARE NOT LIABLE FOR ANY INDIRECT, INCIDENTAL, SPECIAL, CONSEQUENTIAL OR PUNITIVE DAMAGES, OR FOR ANY LOSS OF PROFITS, DATA OR GOODWILL, ARISING FROM YOUR USE OF OR INABILITY TO USE THE SERVICE. Nothing in these terms limits liability that cannot be limited by law. You agree to indemnify Kunkafa against claims arising from your use of the Service or your breach of these terms. 10. Ending the Service You can stop using the Service at any time and delete your account. We may suspend or end access for breach of these terms, and we may change or discontinue parts of the Service; where we can, we give notice. 11. Governing law These terms are governed by the laws of the jurisdiction in which Kunkafa Limited is registered, and disputes go to its courts, unless the law where you live gives you rights that cannot be taken away. Our failure to enforce a right is not a waiver of it. 12. Changes to these terms We may update these terms. The date at the top is the date of the current version; continued use after a change means you accept it. --- ## Privacy Privacy Policy – Kunkafa Privacy Privacy Policy What we collect, why, and what we never do with it. Last updated 7 September 2026. In short We collect what the Service needs to work and, only if you accept the cookie notice, how the site and app perform. Nothing is used for advertising, and nothing is shared with advertisers. 1. What we collect Account: your email address and sign-in method (email and password, Google, or an email link), handled by Firebase Authentication. Plan and billing: your plan and its status. Stripe processes payments and holds card details; we never see your card number. What you do in the app: the markets you follow, the outlooks you save, the alerts and automations you create, and the share cards you publish. Notifications: a device push token, only if you allow notifications. Performance and errors (after consent): page and screen performance, crashes, and how the site is used, through Google Analytics for Firebase and Firebase Performance Monitoring. 2. What we do not collect No advertising identifiers, no advertising cookies, no sale or sharing of personal data with advertisers, and no tracking across other sites. 3. Cookies and identifiers Before you accept the notice, the site sets nothing. If you accept, it stores your choice (a first-party cookie and a browser setting), and Google Analytics sets its client identifier; advertising features are switched off. If you decline, we send you to the home page and set nothing. Signing in to the app uses a session identifier so you stay signed in; the app also keeps your preferences on your device. 4. Who processes data for us Google Firebase (authentication, database, analytics, performance and crash reporting) Stripe (payments and billing) Expo (push notification delivery) The exchange you connect, if you use automation on Pro 5. How long we keep it Account data for as long as your account exists; deleting your account from Settings removes it. Billing records for as long as the law requires. Analytics data for the retention period set in our Google Analytics property. 6. Your rights You can ask to see, correct or delete your personal data, or object to how we use it. Write to contact@kunkafa.com . Where the law where you live gives you further rights, they apply. 7. Children The Service is not for anyone under 18. 8. Changes The date at the top is the date of the current version. We change this page when what we collect changes.