Superteam Earn
Core pillar. Filter AGENT_ALLOWED + low-submission listings. $50–90/op-hr.
Field survey · July 2026 · zero-contact / AI-leveraged
Every market where a solution ships and the money follows — a bounty, a brief, a contest, a court release. 160 of them, mapped and triaged. The ground is mostly dry: about a dozen are worth your hours. Most of the rest fail on one of three faults — the payout is really a lottery, the platform bans AI in substance, or a field of entrenched pros already owns it.
The strike list
Everything that survived verification — the verified pillars plus the alive-and-paying new finds. These are the 12 markets that earn actual hours; the atlas below is the full terrain they sit in.
Core pillar. Filter AGENT_ALLOWED + low-submission listings. $50–90/op-hr.
Fastest cash. 50% upfront over $20, check Universal Scammer List.
Small but low competition; Stripe payout on merge.
Twice-weekly payout post-KYC; ODHack weeks concentrate paid issues.
Keep asks <$5k. Needs 2–3 shipped bounties as proof-of-work. $60–120/op-hr.
AI-native, sponsored model credits, fully async. Squared payout = lottery, but best structural fit in the sweep.
Real lab-funded cash. Manual entry only (~50/50 human/AI, not a Claude-90% pipeline).
57 open events (Jul 2026). AI required, not banned. ~2% win rate → lottery layer; pick low-visibility, high prize/submission events.
Federal (no state registration), uncapped fee. Defunct-business research AI does well; one notarized POA. Highest-ceiling new find.
Highest $/case; guru-saturated. Pick permissive states (GA), attorney partner, >90-day lag.
Firm does the outreach; you deliver research — paid on acceptance, no contingency risk.
Winner-take-all vs. repeat pros; Pearl chart tool mandatory (AI leverage → search only). 1–2 calibration entries on obscure contests.
The full atlas — 14 terrains
Every market found, organized by what you deliver. Filter by verdict, AI-leverage, or contact; search by name. Green = pursue, blue = side-bet, amber = monitor, rust = skip.
Prior-verified base income — objective acceptance-triggered bounties.
Core pillar. Filter AGENT_ALLOWED + low-submission listings. $50–90/op-hr.
Fastest cash. 50% upfront over $20, check Universal Scammer List.
Small but low competition; Stripe payout on merge.
Twice-weekly payout post-KYC; ODHack weeks concentrate paid issues.
Keep asks <$5k. Needs 2–3 shipped bounties as proof-of-work. $60–120/op-hr.
Needs ~$9k float fronted before net-30. GPC channel gutted (EO 14222).
Real 6–12mo R&D obligation; AI = research-misconduct exposure (NIH NOT-OD-25-132).
June 2025 Creativity Standards killed templated/AI bulk listings.
Un-AI-leverageable labor; new-seller caps + 21–31 day holds.
Client-run design / naming / art / video / music contests; win or get selected.
Many contests flag no-AI; low new-account win rate.
AI banned for logos / full artwork.
No AI ban, but massive cheap-contest competition.
Finalist splits help; AI policy unclear.
Zero contact, AI names ubiquitous, low win rate until tier rises.
Licensed-IP challenges repeatable; AI down-scored.
AI forbidden, prizes not cash.
Best royalty economics; anti-AI voting community.
AI detector DQs entries + paywall.
AI-banned; use its directory as an index only.
Idea/pitch phases AI-leverageable; production needs gear + calls.
Treatments AI-friendly; animation briefs most feasible.
Bankrupt Feb 2026.
Rare case-by-case AI-music acceptance; confirm gen-tool license.
AI music rejected.
$299/yr fee eats float.
Multi-year; physical build AI can't do.
~1% advance rate; expert craft.
Fees, years-long, tiny acceptance.
Models / forecasts scored on leaderboards or judged; prize on placement.
AI-native, sponsored model credits, fully async. Squared payout = lottery, but best structural fit in the sweep.
Gen-AI allowed, but pro teams dominate big pools; niche comps winnable.
AI-research; ~0 EV for a generalist.
AI-assist standard; entrenched top forecasters cap $/hr.
2026 closed; next ~Jan 2027; AI permitted.
Reliable cash but needs a 6-mo track record first.
Async, no capital risk; months to Master tier.
No capital risk, USDC, recurring quarterly shots.
Requires NMR stake (capital risk); trivial income sub-$500.
Payout lag measured in years.
Smaller fields than Kaggle; generous pools.
Thin fields improve odds; check Africa-only eligibility.
Research-flavored, favors specialists; small fields.
Tiny academic fields (<50); payment reliability varies.
Niche geospatial; thin competition.
Small energy-sector fields; judged write-ups.
Elite field; only ~4 matches/yr.
Chinese/JP banking gates US payout.
Break / exploit models; paid on verified finding.
Real lab-funded cash. Manual entry only (~50/50 human/AI, not a Claude-90% pipeline).
Episodic; catch the next season.
Public since May 2026; few universal jailbreaks ever verified.
Higher bar (real safety impact, not pure jailbreak); $1M pool.
Rewards model-behavior issues.
Covers model behavior.
Security-impact only; jailbreaks out of scope.
Purpose-built for the jailbreak / model-behavior class.
Small cash, recurring, fellowship pathway.
No money; skill-building only.
Online quals remote; final in-person (Dubai).
Ship an AI-built app/agent to a sponsor brief; judged payout.
57 open events (Jul 2026). AI required, not banned. ~2% win rate → lottery layer; pick low-visibility, high prize/submission events.
AI-builder-native sister ecosystem; ≤90-day payout.
Emerging agent-league category.
Crypto Devpost; AI submissions normal.
Requires a real revenue business + LA finale (Sept 25 2026).
Research / scraping / micro-task deliverables; paid on acceptance, selection, or merge.
Winner-take-all vs. repeat pros; Pearl chart tool mandatory (AI leverage → search only). 1–2 calibration entries on obscure contests.
PATROLL competitor; verify current study volume.
Dormant; ideal when open — standing watch.
Healthiest on-chain DAO board; DOT volatility.
Micro-tickets, instant USDC; optional escrow = nonpayment risk.
Bot competition; sub-$20 tickets.
Marketing dust.
Landing page is a shell; skip pending proof.
Ethical paid fact-check; AI-tool-friendly training.
Raw MT/AI banned; English-only operator.
Dead Sept 2023 (category prototype).
Dead; migrated to Contra.
Standing open calls; paid on editorial acceptance or contest win. AI mostly banned here.
High rate; needs heavy human curation, low acceptance.
Shrunk to Oatmeal ($100, AI-banned); Calypso/Viabella dead.
AI-assisted construction discouraged; elite craft.
Software-assist normal; slow queues.
Fast, byline value; AI-flat gets rejected.
Explicit AI ban; lottery.
Op-eds AI-leverageable; personal essays need lived material.
Must cook + photograph; AI does concept only.
Discontinued Apr 9 2026.
Slow royalty tail; saturated.
Shut down Dec 2025.
Hands-on CAD craft; prizes skew non-cash.
Sponsor / foundation / agency competitions; judged selection.
Alive + paying monthly, but prizes go to repeat experts; ~$2–5/op-hr.
Small purses, many entrants, cheap AI entries.
Smaller fields; social-impact sponsors.
Shrunk but pays; NASA matches recur.
Thinner fields = better odds on open comps.
Rare adult-open big-purse essay; check AI rules.
Thin competition; disclosed AI OK. Watch the 'bounties' tag.
2026 AI-assisted solves landing; low $/hr but historic.
Hardware costs; in-person.
Multi-year team relationship.
IP-scouting funnel.
Funds ventures; cohort program.
Purses are pitch-gated vendor scouting (prior research).
Prior-verified / killed; Challenge.gov sunset Mar 2026.
Refer candidates / intros / leads or post promo clips; paid on conversion.
View lottery, ~$0 month one; filter out crypto-gambling briefs.
Pivoted to B2B SaaS ~Nov 2025; referral board dead.
Hot if admitted; recruiter vetting gate.
Lighter vetting; verify 2026 activity.
Network-bound; zero post-intro mgmt.
Network-bound; deal-close lag.
Owner prospecting + 6–12mo lag + broker-dealer risk.
Legal if referral-only; verify state rules.
Face-on-camera requirement; AI ~10%.
Needs an established creator account.
Assets / mods / experiences accepted into a store or contest; royalties or prizes.
Highest AI-leverage — Claude writes Forge/Fabric mods. Slow build.
AI-embraced; discovery bottleneck; low median.
Luau via Claude; recurring official contests.
Verse via Claude; winner-take-most discovery.
Best mechanic (acceptance → royalty); needs entity + portfolio.
Meta subsidizing hard; gen-AI welcomed; low competition.
Papyrus via Claude; acceptance gate.
No acceptance gate; AI-assist OK except voice cloning.
Launched Jan 2026; early-mover; AI policy unknown.
Elite 3D art; AI-hostile community.
No acceptance gate; declining economy; 2026 fee hikes.
Find / report bugs on real devices; paid on customer acceptance.
Officially provides an AI bug-report tool; human must operate device. $50+/hr on critical hunting.
Largest community; rating ladder gates invites.
Race-to-claim; sporadic cycles.
Fixed report fee decent; thin US invites.
Flat per-test; selection lottery.
Voice narration = AI can't do it.
Perfect model but a ghost town (~7 clients).
Hourly-capped; scheduled windows.
Recorded human think-aloud; no AI leverage (prior research).
Your genuine self in interviews / surveys / studies; per-session pay. AI leverage ~zero — dead-minute filler only.
Real + paying, ~5% acceptance; AI answers bannable. Screener-apply in dead minutes.
~3.4% acceptance; some non-payment reports; gift cards.
40-yr firm; qualification lottery; webcam + recruiter calls.
Async diary; authentic selfie-video.
Needs a verifiable B2B role; polices AI.
Async mock juror; low-volume filler.
Moderated tier clears ~$60/hr.
Fast payer; modest volume.
Paid while asleep; site-gated.
Biggest single payout; run AI portfolio during confinement. In-person + medical risk.
Only if a genuine condition match.
~$10/hr.
AI use = ban; ~$10–15/hr honest (prior research).
In-person single visits (photos / inspections / serving); paid on report acceptance. AI ~zero.
Only ~$50/hr with metro clustering + streetcred.
$50/15min plausible but unreliable.
Async except filming; viral upside.
$50/hr in dense zones; falling rates; door-knocking.
~$50–65/hr; $200–500 startup near float cap.
~$10–20/hr after drive time.
High mileage, low pay.
Research owners/heirs of unclaimed funds; contingency or per-case fee. AI-friendly research; closings need one-off contact.
Federal (no state registration), uncapped fee. Defunct-business research AI does well; one notarized POA. Highest-ceiling new find.
Highest $/case; guru-saturated. Pick permissive states (GA), attorney partner, >90-day lag.
Firm does the outreach; you deliver research — paid on acceptance, no contingency risk.
AI genealogy fit but 6–24mo lag + heir outreach.
One-off disclosure letters OK; screen PI-license states.
Less competed; TX PI-license trap.
Take research-only roles; reject cold-calling agent roles.
Aged pool; can't reach $50/hr.
Debt collection; FDCPA/licensing; capital risk.
Serve only via the subcontract route.
// no markets match those filters
What separates winners from losers
Five structural tells, drawn across the whole universe. Use them to judge any market that isn't on this map yet.
The reliable earners pay on a near-deterministic event — a merge, a validated bug, a court release. The losers dress up as bounties but pay on a subjective judged win, collapsing to lottery odds where 90–98% of entries earn nothing regardless of quality.
The only true fits are markets where AI use is the point or explicitly sanctioned. A large, seductive slice — every paid research study, all UGC video, mock juries, greeting-card and fiction markets — bans AI in substance, because the product is authentic human judgment. Those contribute zero to a Claude-90% strategy no matter how legit the payout.
Thin fields are winnable; giant open pools are not. And where entrenched repeat professionals dominate — PATROLL's top searchers, superforecasters, skin artists — a newcomer's odds crater even inside a “small” field.
Lab-funded, platform-administered, or government-backed channels have near-zero non-payment risk. Optional-escrow or reputation-dependent channels carry real stiff-risk that no amount of AI leverage offsets.
Only Metaculus bot tournaments cleanly hit all four — and even it is gutted by a top-heavy payout curve. Absent that, the durable strategy is the acceptance-bounty pillars for base income, with red-team arenas, hackathons, and surplus/estate research as objective-or-lumpy upside layers on top.