YeNo · Coming soon · proposed rules

Build a bot that buys and sells.
Keep it trading for 12 hours.

Build a bot for YeNo’s five-minute Bitcoin Up/Down markets. It buys a position, sells it, then uses the money left to trade again. Keep the loss on each completed trade small so the bot can keep trading throughout the 12-hour run.

Your deposit: $0
Bot run time: 12 hours
Proposed personal prize: Up to $1,000
Entry dates: To be announced

The 12 hours is the bot’s run time, not your build deadline. Sponsor approval, funding and dates are still pending. You will not need a YeNo account or your own trading money.

The task

What to build

Build a bot for YeNo’s five-minute Bitcoin Up/Down markets. Given market prices and your account information, it decides whether to buy, sell its position or wait. After selling, it uses the remaining money to trade again.

Keep losses small enough to continue for 12 hours. Your bot must complete both the buy and the sale; buying both Up and Down is not required.

Submission

What to submit

Your program implements POST /decide. It receives market and account information, then returns BUY, EXIT or HOLD. The free starter API, real-time market data, baseline bots and validator will ship before entries open.

Submit a repository and commit hash. You can revise up to four times before the cutoff; the selected code stays unchanged from the test API checks through the live final.

Register interest above to receive the starter kit, submission dates and final rules. Builderr operates the funded finalist accounts, credentials, trade execution and safety controls.

Test and final

How we test it

01Build with the test API
The test API uses real-time YeNo market data. Build a bot that buys, sells, and tracks how much money each completed trade keeps or loses. You do not need to fund a trading account.
02Show that your bot works
We run your submitted bot against the test API using real-time data. It must complete its trades, stay within the limits below, and keep working without manual help. Only bots that pass can advance.
03Compete in the real market
Builderr runs the five selected bots, with unchanged code, for the same 12-hour period on separate sponsor-funded accounts. These are real trades. The bot with the most counted trading volume wins.

Ranking

How you score

Complete both parts of a trade: buy a position, then sell it. Choose Up or Down; you do not need to buy both outcomes. Reuse the money from the sale for your next trade. Only one completed buy-and-sell cycle per market counts.

When testing your bot, focus on the loss per completed trade. Count fees and the difference between your buy and sell prices. Smaller losses leave more money for later trades. There are no separate points for speed.

Why small losses matter

Buy for $20 and sell for $19.40: you lose $0.60 before any additional fees. Sell for $19.90 instead: you lose only $0.10. You keep $0.50 more to use on later trades.

We rank bots by total counted trading volume over 12 hours. Low losses help you do more trading; the lowest loss alone does not win. A bot that never trades earns no volume.

How each trade adds to your score

For each completed buy and sale, take the smaller dollar amount and double it. Add those amounts across the run. We call this eligible volume.

Buy: $20.00
Sell: $19.40
Adds to score: $38.80

2 × $19.40 = $38.80 eligible volume

We measure completed trades and your remaining balance using YeNo’s records. Your bot’s reported totals do not determine its score.

Among bots that pass every requirement, the highest eligible volume in the common 12-hour live final wins. Ties go to the higher ending balance, then the smaller maximum drop from an earlier account high.

Which trades count?

Only one completed buy-and-sell cycle per market can score. Rejected or unfilled orders, self-trades, trades between entrants, duplicate cycles, redemptions and unfinished positions do not count. Quotes and submitted order counts add nothing.

Minimum requirements

The bot must complete the test safely.

Reach at least $500 eligible volume in the 12-hour test API run, starting with a $100 test balance.
Finish with at least $50, measured from the platform’s settled cash and inventory records.
At least 95% valid decisions and completion of the scheduled session.
Hold at most one position at a time. Finish with no holdings, pending orders, actions or refunds.
Each buy is capped at $20. The evaluator enforces the $50 floor and allowed markets.
Every counted trade must be independently matched with an external participant and attributable to the submitted bot.
The exact submitted code must run reproducibly in the test API and the live final.

Proposed rewards

Up to $1,000 for one builder.

Each finalist earns 10% of its official live-final eligible volume, capped at $400. First place also receives $600, second $400 and third $200.

Example: $3,000 in eligible volume earns a $300 bonus. Finish second and add $400: $700 total.

The maximum across all finalists is $3,200: up to $2,000 in volume bonuses plus the $1,200 podium pool. Only the single synchronized live final counts for rewards. Practice and qualification runs do not.

YeNo may also invite builders to work on its future bot marketplace. The rewards, YeNo Bot Expert credential and any marketplace invitation still need approval before entries open.

Before entries open

The remaining rules will be published first.

We still need to define what counts as a valid decision, how the top five are selected, the test API run schedule, the live final’s starting balance, how the $50 floor is measured, and what happens after an outage or failed run. Payout eligibility and replacement rules also need confirmation. These details will be fixed before you are asked to submit.

Technical API, market rules and live-test notes

Build against this

The input and output format.

The free starter server, real-time market data, baseline bots, and validator will ship before entries open.

POST /decideJSON in · JSON out

// evaluator → your bot
{
  "market": {
    "id": "opaque",
    "seconds_left": 214,
    "yes": { "bid": 0.54, "ask": 0.55 },
    "no":  { "bid": 0.44, "ask": 0.45 }
  },
  "portfolio": { "cash_usd": 83.40, "positions": [] },
  "limits": { "max_buy_usd": 20, "max_open_positions": 1 }
}

// your bot → evaluator
{ "action": "BUY", "outcome": "YES", "max_cost_usd": 12 }

You build the bot. You do not fund or operate a trading account.

Build and qualify using the test API and real-time data. If your bot passes the checks and is selected for the top five, Builderr runs that exact code with sponsor funds and the published limits. Its accepted orders then become real YeNo market trades.

HOLD

Skip a weak, thin, stale, or late setup.

BUY

Choose YES or NO and a maximum dollar cost.

EXIT

Request a full exit; the evaluator reconciles shares.

Submit a repository and commit hash. You can revise up to four times before the cutoff; the selected commit is immutable from the test API checks through the live final.

Market scope

Only Bitcoin five-minute Up/Down markets.

The same market type applies to every bot.

Allowed live market

BTC five-minute Up/Down

The evaluator resolves each current market ID. Bots cannot hardcode IDs or switch market families. A 12-hour window contains up to 144 eligible market cycles.

Why this scope

Our executor, rollover logic, reference feed, test API, and live test history are BTC-specific. The $20 canary passed; $10–$15 is the recommended working range when the book is thinner.

Field notes

Handling failed or incomplete trades.

The hard part is not returning BUY. It is staying correct when a venue response is late, partial, stale, or ambiguous.

Check whether an order landed before retrying

After a timeout, check YeNo’s records first. Retrying blindly can create a second position.

A rejection is not a trade

Failed, unmatched, refunded, and zero-share attempts earn no action or volume credit.

Sell the shares the account actually holds

Partial fills and rounding can change the position size. Check YeNo’s records before every exit and finish with no holdings.

Do not enter too late

Skip stale prices and markets that are about to close. The bot must also handle market rollover and settlement.

Keep enough money to continue

A bot that loses too much early cannot finish the window. Trade smaller or wait when another trade would put the remaining balance at risk.

Only completed trades count

Submitted orders, displayed prices and retries add nothing. A buy and its matched sale must both complete before they add eligible volume.

$20 test trade · completed Sep 9, 2026

One $20 round trip worked. Start smaller when the market is thin.

This one trade showed that a complete buy and sale was possible. It does not show that every $20 trade will fill at a similar price.

YeNo's current guide states $4–$50 per order. Our guarded live test bought $20 of YES for 20.5729 shares and exited the full position 4.3 seconds later for $19.28. It created $39.28 in gross matched volume ($38.56 eligible volume under the competition formula) and finished with zero positions or pending actions. The trade cashflow showed a $0.72 loss; authoritative promo balance moved from $51.73 to $50.98, so scoring uses settled balance and inventory—not a bot's local estimate.

Competition rule: $20 is the hard maximum, not the target. The starter bots will default to $10–$15. Larger orders can walk a thin book, increase exit slippage, and consume the $50 loss allowance faster. A bot should use $20 only when visible depth and loss headroom support the full round trip.

Remaining launch dependency: YeNo's written approval for automated finalist traffic and the final sponsor-funded reward terms.