Best AI Model Leaderboard
Real, live odds each AI lab has the single best model — every tracked lab ranked on the same real question, for every real period Polymarket is currently running this contract family for.
Based on real, live prediction-market prices, Anthropic is favored at 99.00% to have the single best AI model by the end of August 2026, 98.4pt ahead of OpenAI — drawn from 570 live AI markets tracking $559.6K in real 24h volume.
Overall AI Leader Recognition
every real tracked period, volume & recency-weightedHow this score is calculated
Not just the nearest resolution date — every real period Polymarket runs this "single best AI model" contract for, combined with a real volume-weighted average that also discounts each period by how far out it resolves (a period backed by more real trading pulls the score toward it more, but a heavily-traded contract months away still can't fully outvote a thinly-traded one resolving imminently — the score stays anchored to who's actually favored right now). A lab genuinely closing on the leader fast — a real, rising 1-week change in the soonest tracked period — also gets a small, capped boost toward the leader before the blend runs, so a fast-moving real challenger doesn't score as far behind as a flat one sitting at the same raw probability; it can never fabricate a lead or overtake the real leader on trend alone. The period-by-period table below shows the same underlying real contracts broken out one date at a time, unadjusted.
Ranked By August 2026
9 labs trackedHow Odds Differ By Month
4 real periods trackedToday's real live odds for the single best AI model contract, across every real resolution month Polymarket is currently running it for — not a historical trend, a side-by-side read of whether each lab's lead holds up further out or fades.
How The Race Is Changing
Not enough archived history yet for a trend chart — 5 real days recorded, 7 required. Check back as more days record.
Lead Changes
Forecast Accuracy — Scored Before the Market Closed
Every figure here is scored against the price the market was genuinely showing a set time before it resolved — recovered from Polymarket's own order-book price history, not from the settled closing price. A settled market already knows the answer, so scoring it measures nothing.
The accuracy archive couldn't be reached just now. Nothing is estimated in its place — reload to try again.
Every Live AI Market, Organized
AGI & Singularity Timing1 market · $282 24h volFull Page →▼
Model Rankings (Best AI Model/Lab)283 markets · $275.3K 24h volFull Page →▼
Overall Best AI Model
121+ 116 more
Chinese AI Models
31+ 26 more
Coding AI Labs
39+ 34 more
Math AI Labs
43+ 38 more
AI Agents
29+ 24 more
Text-to-Image AI
12+ 7 more
Other Rankings
8+ 3 more
Model Releases60 markets · $175.6K 24h volFull Page →▼
+ 21 more
Other
14+ 9 more
OpenAI
11+ 6 more
Anthropic
4Safe Superintelligence
3Benchmark & Leaderboard Scores76 markets · $32.0K 24h volFull Page →▼
Arena Leaderboard
3Humanity's Last Exam
15+ 10 more
LiveBench
44+ 39 more
Other Benchmarks
14+ 9 more
Compute & Hardware Pricing50 markets · $28.9K 24h volFull Page →▼
+ 42 more real markets
Company Valuation & IPO50 markets · $32.5K 24h volFull Page →▼
Anthropic
26+ 21 more
OpenAI
18+ 13 more
DeepSeek
3Other AI Markets50 markets · $15.0K 24h vol▼
Anthropic
24+ 19 more
Other
21+ 16 more
AI News
Real Hacker News stories mentioning "AI"
- Don't post generated/AI-edited comments. HN is for conversation between humans171d ago
- Open source AI is the path forward767d ago
- My AI skeptic friends are all nuts453d ago
- An AI agent published a hit piece on me198d ago
- Gemini AI997d ago
- I believe there are entire companies right now under AI psychosis106d ago
Frequently Asked Questions
Which AI company is #1 right now?
Based on real, live prediction-market prices, Anthropic is currently favored at 99.00% to have the single best AI model by the end of August 2026.
Why does the leaderboard change by period?
Polymarket runs this exact "best AI model" contract for several real resolution dates at once — a lab favored to lead by the nearest date isn't necessarily favored to still be ahead months later, so each period gets its own real, independently-ranked leaderboard rather than one blended number.
How are the other AI markets grouped?
By theme (model-ranking contracts, new-model releases, benchmark/leaderboard scores, compute/hardware pricing, and company valuation/IPO markets) or by which lab a question names — both derived automatically from each market's real question text, not manually curated.
Does this measure app usage or downloads?
No — no platform this site tracks publishes usage-share data. This reflects real money traders are staking on which lab's model will actually be judged best, not popularity.





