Benchmark Almanac

fev-bench

ci-proto 90.0 tops its leaderboard
Not in model cards

Models forecast time series across 100 tasks, including tasks with additional variables available to the forecaster.

The evaluation reports a skill score and win rate across tasks.

Made by
AutoGluon team
Released
September 2025
Kind
Real-world events
Level
University
Scored
Checked automatically
Answers
Prediction
Tasks
All public
Measures
Capability
Subject
Mathematics
Language
English
In and out
Time series → Time series
Interaction
Single turn
Refresh
One fixed set
Run by
Anyone
35
Entries on its own leaderboard, 0 models we track

ci-proto tops the leaderboard with 90.0

Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders.

  1. 1 ci-proto SQL 90.0
  2. 2 TimeH2O-v1 SQL 89.6
  3. 3 Timestral-209M SQL 87.1
  4. 4 t0-gamma SQL 84.7
  5. 5 Simons-297M SQL 82.0
  6. 6 TimesFM-3 SQL 79.3
  7. 7 iAmTime-base-forecast SQL 73.7
  8. 8 Chronos-2 SQL 72.2
  9. 9 t0-beta SQL 69.9
  10. 10 Toto-2.0-2.5B SQL 69.4
  11. 11 TiRex-2 SQL 69.2
  12. 12 TiRex SQL 62.2
  13. 13 Toto-2.0-22m SQL 61.6
  14. 14 TimesFM-2.5 SQL 58.4
  15. 15 TabPFN-TS-3 SQL 55.5
  16. 16 TS-ICL SQL 55.1
  17. 17 CITRAS-FM SQL 52.9
  18. 18 FlowState SQL 52.6
  19. 19 Toto-1.0 SQL 51.6
  20. 20 TabPFN-TS SQL 50.5
  21. 21 Moirai-2.0 SQL 46.2
  22. 22 Chronos-Bolt SQL 45.4
  23. 23 TFT SQL 34.7
  24. 24 Sundial-Base SQL 30.8
  25. 25 Stat. Ensemble SQL 30.6
  26. 26 PatchTST SQL 30.0
  27. 27 DeepAR SQL 28.9
  28. 28 AutoARIMA SQL 26.6
  29. 29 AutoETS SQL 22.5
  30. 30 CatBoost SQL 21.8
  31. 31 LightGBM SQL 20.3
  32. 32 AutoTheta SQL 17.9
  33. 33 Seasonal Naive SQL 12.0
  34. 34 Naive SQL 7.80
  35. 35 Drift SQL 7.00

Source: Official fev-bench leaderboard

#EntryScoreSettingDateLinkSource
1 ci-proto 90.0 SQL huggingface.co Official leaderboard
2 TimeH2O-v1 89.6 SQL huggingface.co Official leaderboard
3 Timestral-209M 87.1 SQL huggingface.co Official leaderboard
4 t0-gamma 84.7 SQL huggingface.co Official leaderboard
5 Simons-297M 82.0 SQL huggingface.co Official leaderboard
6 TimesFM-3 79.3 SQL huggingface.co Official leaderboard
7 iAmTime-base-forecast 73.7 SQL huggingface.co Official leaderboard
8 Chronos-2 72.2 SQL huggingface.co Official leaderboard
9 t0-beta 69.9 SQL huggingface.co Official leaderboard
10 Toto-2.0-2.5B 69.4 SQL huggingface.co Official leaderboard
11 TiRex-2 69.2 SQL huggingface.co Official leaderboard
12 TiRex 62.2 SQL huggingface.co Official leaderboard
13 Toto-2.0-22m 61.6 SQL huggingface.co Official leaderboard
14 TimesFM-2.5 58.4 SQL huggingface.co Official leaderboard
15 TabPFN-TS-3 55.5 SQL huggingface.co Official leaderboard
16 TS-ICL 55.1 SQL huggingface.co Official leaderboard
17 CITRAS-FM 52.9 SQL huggingface.co Official leaderboard
18 FlowState 52.6 SQL huggingface.co Official leaderboard
19 Toto-1.0 51.6 SQL huggingface.co Official leaderboard
20 TabPFN-TS 50.5 SQL huggingface.co Official leaderboard
21 Moirai-2.0 46.2 SQL huggingface.co Official leaderboard
22 Chronos-Bolt 45.4 SQL huggingface.co Official leaderboard
23 TFT 34.7 SQL github.com Official leaderboard
24 Sundial-Base 30.8 SQL huggingface.co Official leaderboard
25 Stat. Ensemble 30.6 SQL nixtlaverse.nixtla.io Official leaderboard
26 PatchTST 30.0 SQL github.com Official leaderboard
27 DeepAR 28.9 SQL github.com Official leaderboard
28 AutoARIMA 26.6 SQL nixtlaverse.nixtla.io Official leaderboard
29 AutoETS 22.5 SQL nixtlaverse.nixtla.io Official leaderboard
30 CatBoost 21.8 SQL nixtlaverse.nixtla.io Official leaderboard
31 LightGBM 20.3 SQL nixtlaverse.nixtla.io Official leaderboard
32 AutoTheta 17.9 SQL nixtlaverse.nixtla.io Official leaderboard
33 Seasonal Naive 12.0 SQL nixtlaverse.nixtla.io Official leaderboard
34 Naive 7.80 SQL nixtlaverse.nixtla.io Official leaderboard
35 Drift 7.00 SQL nixtlaverse.nixtla.io Official leaderboard

No lab model card reports it yet

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.

More in Forecasting

All 47 in Forecasting
  • ForecastBench In use Models give probabilities for 500 auto-generated questions per round about future events, drawn from prediction markets and time-series datasets. Forecasts are scored with difficulty-adjusted Brier scores once the questions resolve. GPT-5.5 0.11
  • Prophet Arena In use Models and agents forecast open Kalshi prediction-market events and are scored once the events resolve. The live board ranks a difficulty-adjusted Arena score and also reports calibration error and market return. Claude Opus 5.5 86.2
  • Guesswork (The Aggregate) Not in model cards Models predict benchmark scores before publication. The maker ranks forecasts by mean absolute error after scaling each error by the variation in that benchmark’s scores. GPT-6 Astra 0.56
  • AGI-Pilled Bench Not in model cards Models forecast when AI will reach milestones that already happened in 2026. The score is the median years after 2026 across the forecast tasks. Claude Fable 5.2 6.50
  • Dutch Book Bench Not in model cards Models assign probabilities to related events in separate conversations. The grader checks whether inconsistent probabilities let a bettor guarantee a profit. Gemini 3.8 Flash 89.2%
  • Echo General AI Prediction Leaderboard Not in model cards Models forecast unresolved events drawn from prediction markets, synthesized news questions and expert-written questions. Forecasts made at aligned times are compared in pairwise battles, and an Elo fit ranks the models. DeepSeek-V4-Pro 1,018

Data audit

All 8 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 2 pass 3 notes
  • Pass Independent runs Scores from the official leaderboard (35 entries).
  • Pass Models scored 35 entries on its leaderboard, none a model we track.
  • Note Newest score No scores yet.
  • Note Frontier models No models scored yet.
  • Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on huggingface.co: 35 entries at the last read, none a model we track.
  • Pass Last read November 20th, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check