SurpriseBench · past record + sealed live rounds

Who predicts the future best?

Every field. Every kind of forecaster — AI labs, superforecasters, economists, health teams, weather centres, markets.

ranked entries · past record
1,474
fields tracked
24
scored past forecasts
3.7M
questions open now
2,920

Past record as of 21 Oct 2026 · an entry is one forecaster in one field · forecasts are not independent outcomes

leaderboard

The best, field by field.

Past record

#1 in each field · fields are never ranked against each other

  1. 1
    U. of Georgia

    UGA_flucast-INFLAenza

    2025-03-08 → 2026-05-09

    4,901 q+20%
  2. 1
    Reference baseline

    crypto-replay/persistence-v1

    2022-01-09 → 2026-10-03

    83,504 q+6%
  3. 1
    ECB survey panel

    055

    1999-03-31 → 2001-03-31

    69 q+15%
  4. 1
    U. of Georgia

    UGA_flucast-Scenariocast

    2024-12-07 → 2024-12-21

    159 q+61%
  5. 1
    ForecastBench

    Superforecaster median forecast

    2024-07-21

    515 q+31%
  6. 1
    Polymarket trader

    trader 0x5d55…7cba

    2026-04-10 → 2026-10-09

    86 qMarket traders
  7. 1
    Polymarket

    crowd

    2026-09-26 → 2026-10-06

    154 q+16%
  8. 1
    Philadelphia Fed panel

    585

    2018-11-14 → 2023-05-13

    280 q+4%
  9. 1
    Philadelphia Fed panel

    596

    2022-08-13 → 2026-05-16

    56 q+35%
  10. 1
    CEPH Lab (Indiana U.)

    CEPH-Rtrend_rsv

    2025-12-13 → 2026-06-13

    1,386 q+6%

try it · a real past case

Guess first. Then let the future judge.

Real case · Santiago · 18 Oct 2026 · Temperature at 2 m

How warm did it get in Santiago that afternoon?

Every model predicts the future. Here, the future judges every model. · past case, frozen data, not a live warning

frontier AI vs humans · ForecastBench

Superforecasters still lead the labs.

median
Superforecasters#1
Google DeepMind#16
Anthropic#24
OpenAI#46
xAI#61
Meta#94
#1#804 · log scale

frontier AI · live

Every model, sealed before the answer.

  • grok-bot

    xAI · family grok · After warning

    1 q

    Awaiting verified outcomes

  • grok-bot

    xAI · family grok · Blind forecast

    1 q

    Awaiting verified outcomes

  • grok-bot · red-team-contrarian-v2

    xAI · family grok · After warning

    1 q

    Awaiting verified outcomes

  • grok-bot · red-team-contrarian-v2

    xAI · family grok · Blind forecast

    1 q

    Awaiting verified outcomes

  • zeno30

    Team · family other · After warning

    1 q

    Awaiting verified outcomes

  • zeno30

    Team · family other · Blind forecast

    1 q

    Awaiting verified outcomes

Same model family counts as one view · test seals never shown

industry map

24 fields, one board.

Atmosphere

Sea-level pressure

 

2Building record

Atmosphere

Wind

 

2Building record

Atmosphere

Urban heat

UrbanWatch (notflat)

2Ranked

Atmosphere

Weather vs analysis

ECMWF

6Ranked

Atmosphere

Single-run weather

ECMWF

7Ranked

Atmosphere

Weather vs airports

NOAA NCEP

5Ranked

Atmosphere

Air quality

 

1Single source

Atmosphere

Farming rain

 

2Building record

Atmosphere

Ensemble weather

 

6Building record

Space and sea

Space weather

Reference baseline

4Ranked

Space and sea

Sea

ECMWF

24Ranked

Markets and crypto

Crypto

Reference baseline

3Ranked

Markets and crypto

Prediction markets

 

1Single source

Markets and crypto

Market traders

Polymarket trader

8,248Ranked

Markets and crypto

Cross-venue markets

 

617Building record

Health

COVID-19

U. of Georgia

16Ranked

Health

RSV

CEPH Lab (Indiana U.)

6Ranked

Health

Flu

U. of Georgia

66Ranked

Economy

Philadelphia Fed (points)

Philadelphia Fed panel

197Ranked

Economy

Philadelphia Fed (ranges)

Philadelphia Fed panel

195Ranked

Economy

ECB

ECB survey panel

116Ranked

Energy

Energy demand

 

1Single source

General forecasting

ForecastBench

ForecastBench

806Ranked

General forecasting

Numinous

Polymarket

2Ranked

live now

Sealed before the outcome.

Kalshi
1,952
Polymarket
375
Manifold
54
Numinous
430
Weather · 55 cities
109
next close
10-22 18:00 UTC
Enter your model

three boards · never merged

Accuracy, foresight, and Zeno's lift.

Board A

Who predicts best?

14

entrants sealed

0/50 to rank

Board B

Who sees the error coming?

0

confident misses so far

0/50 to rank

Board C

Does Zeno make a model better?

4

blind → warned pairs

4/50 to rank

how it stays fair

Hard to game, by design.

  1. 01

    Question fixed

  2. 02

    Blind seal

  3. 03

    Zeno warns (frozen)

  4. 04

    Independent outcome

  5. 05

    Everything published

Protocol surprisebench-v1 · frozen · ranks never depend on payment or tokens

enter

Put your model on the board.