Skip to main content

WeirdML v2 (WeirdML)

A machine-learning engineering benchmark that tests whether LLMs can train models on novel datasets, write PyTorch code, and improve through iterative feedback.

How BenchLM shows WeirdML

BenchLM mirrors the top public WeirdML v2 rows from the official CSV. WeirdML tests whether models can do machine-learning engineering on 17 novel datasets, reporting average accuracy across tasks.

WeirdML is display only on BenchLM because the task is a specialized ML-agent evaluation with iterative feedback and execution constraints. BenchLM stores it as context rather than a weighted model-only benchmark.

25 mirrored top rows130 official CSV rows17 ML tasksOfficial CSVDisplay only

Average accuracy on WeirdML — WeirdML v2

BenchLM mirrors the published average accuracy view for WeirdML. gpt-5.5 (xhigh) leads the public snapshot at 84.91% , followed by gpt-5.5 (high) (83.90%) and claude-opus-4.8 (xhigh) (82.89%). BenchLM does not use these results to rank models overall.

25 modelsExternal benchmark mirrorsCurrentDisplay onlyUpdated WeirdML v2

The published WeirdML snapshot is tightly clustered at the top: gpt-5.5 (xhigh) sits at 84.91%, while the third row is only 2.02 points behind. The broader top-10 spread is 9.46 points, so many of the published scores sit in a relatively narrow band.

25 models have been evaluated on WeirdML. The benchmark falls in the External benchmark mirrors category. BenchLM tracks this category separately from its weighted global scoring system, so these results are best compared on the dedicated Korean benchmark views. WeirdML is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.

About WeirdML

Year

2026

Tasks

17 novel ML engineering tasks

Format

Average accuracy across tasks

Difficulty

Novel dataset modeling and iterative debugging

WeirdML v2 evaluates models on 17 unusual ML tasks and reports average accuracy across tasks from the official CSV. BenchLM mirrors the top official rows as display-only ML-agent evidence.

BenchLM freshness & provenance

Version

WeirdML 2026

Refresh cadence

Quarterly

Staleness state

Current

Question availability

Public benchmark set

CurrentDisplay only

BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.

Average accuracy table (25 models)

1
gpt-5.5 (xhigh)gpt-5.5:xhigh
84.91%
2
83.90%
3
claude-opus-4.8 (xhigh)claude-opus-4.8:xhigh
82.89%
4
claude-opus-4.6 (high)claude-opus-4.6:or_adaptive
77.95%
5
77.90%
6
77.70%
7
claude-opus-4.7 (high)claude-opus-4.7:adaptive
76.44%
8
claude-opus-4.7 (no thinking)claude-opus-4.7:no_thinking
76.40%
9
claude-opus-4.8 (medium)claude-opus-4.8:medium
76.04%
10
claude-opus-4.7 (max)claude-opus-4.7:max
75.45%
11
72.19%
12
gemini-3.1-pro-preview (high)gemini-3.1-pro-preview
72.07%
13
claude-opus-4.8 (no thinking)claude-opus-4.8:no_thinking
70.45%
14
gemini-3-pro-preview (high)gemini-3-pro-preview
69.93%
15
gpt-5.5 (no thinking)gpt-5.5:no_thinking
67.15%
16
claude-sonnet-4.6 (medium)claude-sonnet-4.6:or_adaptive
66.07%
17
65.87%
18
63.74%
19
gpt-5.2 (medium)gpt-5.2:medium
63.44%
20
gemini-3.5-flash (high)gemini-3.5-flash
62.64%
21
gemini-3-flash-preview (high)gemini-3-flash-preview:or
61.60%
22
60.77%
23
gpt-5 (high)gpt-5-2025-08-07
60.70%
24
gpt-5-pro (high)gpt-5-pro-2025-10-06
60.39%
25
60.30%

FAQ

What does WeirdML measure?

A machine-learning engineering benchmark that tests whether LLMs can train models on novel datasets, write PyTorch code, and improve through iterative feedback.

Which model leads the published WeirdML snapshot?

gpt-5.5 (xhigh) currently leads the published WeirdML snapshot with 84.91% average accuracy. BenchLM shows this benchmark for display only and does not use it in overall rankings.

How many models are evaluated on WeirdML?

25 AI models are included in BenchLM's mirrored WeirdML snapshot, based on the public leaderboard captured on WeirdML v2.

Last updated: WeirdML v2 · mirrored from the public benchmark leaderboard

The AI models change fast. We track them for you.

For engineers, researchers, and the plain curious — a weekly brief on new models, ranking shifts, and pricing changes.

Free. No spam. Unsubscribe anytime.