Benchmark Almanac

FeatEng

o1-preview 11.1% leads since October 2024
Not in model cards

Models write Python functions to transform 103 tabular datasets. The score measures how much those changes reduce prediction error for an XGBoost model.

The authors publish a ranking of 23 model configurations and an automated machine learning baseline.

Released
October 2024
Kind
Data science and SQL
Level
Expert
Scored
Checked automatically
Answers
Code
Tasks
All public
Measures
Capability
Language
English
In and out
Text, Tables → Code
Interaction
Single turn
Environment
Code sandbox
Refresh
One fixed set
Run by
Anyone
Strongest model tested
o1-mini #142 on the capability index
Last model added
September 2024
Scores refreshed
November 21st, 2026
22
Entries in its paper, 16 models we track

o1-preview leads with 11.1% among models released by September 2024 o1-preview leads with 11.1% among models released by September 2024

Best score per entry from the benchmark's paper, read on November 21st, 2026. A paper’s table doesn’t add newer models. 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. This view shows only models we track, each at its best.

The newest model with a score here came out in September 2024, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 o1-preview 11.1%
  2. 2 Gemini 1.5 Pro Sep 2024 7.77%
  3. 3 Gemini 1.5 Flash Sep 2024 7.33%
  4. 4 o1-mini 7.20%
  5. 5 GPT-4o Aug 2024 7.04%
  6. 6 DeepSeek-V2.5 Sep 2024 6.85%
  7. 7 Claude 3.5 Sonnet 6.75%
  8. 8 Mistral Large 2407 6.46%
  9. 9 GPT-4o May 2024 6.41%
  10. 10 Llama 3.1-405B 6.30%
  11. 11 GPT-4o mini 6.04%
  12. 12 GPT-4 Turbo Apr 2024 5.78%
  13. 13 Codestral 22B 5.76%
  14. 14 Claude 3 Opus 5.02%
  15. 15 Llama 3.1-70B 4.58%
  16. 16 Claude 3 Haiku 4.39%
  17. 17 Mistral Small 2409 4.00%
  18. 18 Mixtral 8x22B v0.1 3.46%
  19. 19 Yi Coder 9B 2.91%
  20. 20 Jamba 1.5 Large 2.36%
  21. 21 Command R+ 08-2024 1.25%
  22. 22 Mixtral 8x7B 1.01%

Source: The FeatEng paper

#EntryScoreSettingDateSource
1 o1-preview 11.1% Sep 2024 Official leaderboard
2 Gemini 1.5 Pro Sep 2024 7.77% Sep 2024 Official leaderboard
3 Gemini 1.5 Flash Sep 2024 7.33% Aug 2024 Official leaderboard
4 o1-mini 7.20% Sep 2024 Official leaderboard
5 GPT-4o Aug 2024 7.04% Aug 2024 Official leaderboard
6 DeepSeek-V2.5 Sep 2024 6.85% Sep 2024 Official leaderboard
7 Claude 3.5 Sonnet 6.75% Jun 2024 Official leaderboard
8 Mistral Large 2407 6.46% Official leaderboard
9 GPT-4o May 2024 6.41% May 2024 Official leaderboard
10 Llama 3.1-405B 6.30% Jul 2024 Official leaderboard
11 GPT-4o mini 6.04% Jul 2024 Official leaderboard
12 GPT-4 Turbo Apr 2024 5.78% Apr 2024 Official leaderboard
13 Codestral 22B 5.76% Official leaderboard
14 Claude 3 Opus 5.02% Feb 2024 Official leaderboard
15 Llama 3.1-70B 4.58% Jul 2024 Official leaderboard
16 Claude 3 Haiku 4.39% Mar 2024 Official leaderboard
17 Mistral Small 2409 4.00% Official leaderboard
18 Mixtral 8x22B v0.1 3.46% Official leaderboard
19 Yi Coder 9B 2.91% Official leaderboard
20 Jamba 1.5 Large 2.36% Aug 2024 Official leaderboard
21 Command R+ 08-2024 1.25% Official leaderboard
22 Mixtral 8x7B 1.01% Dec 2023 Official leaderboard
#ModelScoreSettingDateSource
1 o1-preview 11.1% Sep 2024 Official leaderboard
2 Gemini 1.5 Pro Sep 2024 7.77% Sep 2024 Official leaderboard
3 Gemini 1.5 Flash Sep 2024 7.33% Aug 2024 Official leaderboard
4 o1-mini 7.20% Sep 2024 Official leaderboard
5 GPT-4o Aug 2024 7.04% Aug 2024 Official leaderboard
6 DeepSeek-V2.5 Sep 2024 6.85% Sep 2024 Official leaderboard
7 Claude 3.5 Sonnet 6.75% Jun 2024 Official leaderboard
8 GPT-4o May 2024 6.41% May 2024 Official leaderboard
9 Llama 3.1-405B 6.30% Jul 2024 Official leaderboard
10 GPT-4o mini 6.04% Jul 2024 Official leaderboard
11 GPT-4 Turbo Apr 2024 5.78% Apr 2024 Official leaderboard
12 Claude 3 Opus 5.02% Feb 2024 Official leaderboard
13 Llama 3.1-70B 4.58% Jul 2024 Official leaderboard
14 Claude 3 Haiku 4.39% Mar 2024 Official leaderboard
15 Jamba 1.5 Large 2.36% Aug 2024 Official leaderboard
16 Mixtral 8x7B 1.01% Dec 2023 Official leaderboard

Rankings by part

We read 6 parts from the benchmark’s paper, each ranking models on one part of the test, such as Classification, Regression and Failure. o1-preview leads 5 of them, and Mixtral 8x7B leads the other one. Arrows show how far a model moved from its overall rank among the models we track.

Classification

The paper reports this score on classification datasets. 16 models we track have a score here.

  1. 1 o1-preview 8.90%
  2. 2 Gemini 1.5 Pro Sep 2024 6.18%
  3. 3 o1-mini ↑1 5.91%
  4. 4 Claude 3.5 Sonnet ↑3 5.79%
  5. 5 Gemini 1.5 Flash Sep 2024 ↓2 4.70%
  6. 6 Llama 3.1-405B ↑3 4.43%
  7. 7 DeepSeek-V2.5 Sep 2024 ↓1 4.34%
  8. 8 GPT-4o Aug 2024 ↓3 4.27%
  9. 9 GPT-4 Turbo Apr 2024 ↑2 4.16%
  10. 10 GPT-4o May 2024 ↓2 3.74%

Number one over time

Lab claims count from the day the card came out, independent runs from the model’s release.

OpenAI
Benchmark paper Not tested since Oct 2024
o1-preview 11.1%

Source: The FeatEng paper

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

Plot with other benchmarks
In the benchmark’s paper Best so far
0% 20% 40% 60% 80% 100% 2024 Near the ceiling (85%) o1-preview Claude 3.5 Sonnet Claude 3 Opus 0% 20% 40% 60% 80% 100% 2024 Near the ceiling (85%) o1-preview Mixtral 8x7B

Source: Official FeatEng 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 Coding

All 523 in Coding
  • SWE-bench Verified In use 500 human-validated GitHub issues from 12 Python repos. Agent writes a patch that must pass hidden tests. Claude Opus 4.7 83.5%
  • LiveCodeBench v6 In use Competitive programming problems collected after model cutoffs from LeetCode, AtCoder, Codeforces. Scored pass@1 on hidden tests. DeepSeek-V4-Pro 93.5%
  • SWE-bench Multilingual In use 300 GitHub-issue tasks across nine programming languages in SWE-bench format. Patch graded by repository tests. Claude Fable 5.2 96.6%
  • DeepSWE In use 113 original long-horizon engineering tasks across 91 repos and five languages, written from scratch. Hand-written verifiers accept alternative solutions. GPT-6 Astra 74.1%
  • SciCode In use Scientist-curated research coding problems split into subproblems across natural sciences. Code checked against test cases. Seed2.1 Pro 59.8%
  • Codeforces rating In use Model solves recent Codeforces contest problems. An Elo-style rating is computed from simulated contest performance. DeepSeek V4.1 Flash 3,471

Data audit

9 of 10 checks pass, 1 warning. Look first at the Newest score check. The newest came 2 years ago, in September 2024, for Gemini 1.5 Pro. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 1 warning 3 pass 1 note
  • Pass Independent runs Scores from the benchmark’s paper (22 entries).
  • Pass Models scored 16 models we track, plus 6 other entries on its leaderboard.
  • Warning Newest score The newest came 2 years ago, in September 2024, for Gemini 1.5 Pro.
  • Note Frontier models None of the top 10 has a score. The strongest tested is o1-mini (#142).
  • Pass Headroom The best is 11.1%, under 85%, so it still separates models.
Leaderboard import 3 pass
  • Pass Importer We read the board on arxiv.org: 17 rows for models we track and 6 other entries at the last read.
  • Pass Last read November 21st, 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