Benchmark Almanac

Climate-Change-NER

GPT-4o May 2024 0.44 leads since May 2024
Not in model cards

Models identify and label climate terms in scientific abstracts. The authors provide 13 categories and reserve 75 abstracts for testing.

Elena Volkanovska reported F1 scores for four language models on 177 test sentences.

Made by
IBM Research and NASA
Released
April 2024
Kind
Lab and data work
Level
University
Scored
Checked automatically
Answers
Short answer
Tasks
All public
Measures
Capability
Language
English
In and out
Text → Text
Interaction
Single turn
Items from
Real work and use
Refresh
One fixed set
Run by
Anyone
Strongest model tested
GPT-4o May 2024 #162 on the capability index
Last model added
July 2024
Scores refreshed
November 20th, 2026
4
Models scored in its paper

2 models share the top score of 0.44 among models released by July 2024

Best score per model from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart.

The newest model with a score here came out in July 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 GPT-4o May 2024, 5 random examples, tokens 0.44
  2. 2 Llama 3.1-405B 5 similar examples, strings 0.44
  3. 3 Llama 3.1-70B 3 similar examples, strings 0.40
  4. 4 GPT-4o mini 5 similar examples, strings 0.38

Source: The Climate-Change-NER paper

#ModelScoreSettingDateSource
1 GPT-4o May 2024 0.44 5 random examples, tokens May 2024 Official leaderboard
2 Llama 3.1-405B 0.44 5 similar examples, strings Jul 2024 Official leaderboard
3 Llama 3.1-70B 0.40 3 similar examples, strings Jul 2024 Official leaderboard
4 GPT-4o mini 0.38 5 similar examples, strings Jul 2024 Official leaderboard

Scores at each effort level

Models the leaderboard ran at two or more effort levels. We don’t have enough cost data here to say what the extra effort costs.

Lowest effort to highest, light to dark
GPT-4o
3 random ex… 5 random ex… 3 random ex… 4 similar e… 5 similar e… 5 random ex…
0.44
Llama 3.1-405B
3 similar e… 4 random ex… 3 random ex… 5 similar e…
0.44
Llama 3.1-70B
4 similar e… 5 similar e… 3 random ex… 4 random ex… 5 random ex…
0.40
GPT-4o mini
3 random ex… 3 random ex… 5 random ex… 5 similar e…
0.38

Source: The Climate-Change-NER paper

ModelEffortHarnessScoreCost per taskOutput tokens
GPT-4o May 2024 5 random examples, tokens 0.44
Llama 3.1-405B 5 similar examples, strings 0.44
GPT-4o May 2024 5 similar examples, strings 0.43
Llama 3.1-405B 3 similar examples, strings 0.43
Llama 3.1-405B 4 similar examples, strings 0.43
GPT-4o May 2024 5 similar examples, tokens 0.42
Llama 3.1-405B 3 random examples, tokens 0.42
Llama 3.1-405B 5 random examples, strings 0.42
Llama 3.1-405B 4 similar examples, tokens 0.42
GPT-4o May 2024 4 random examples, tokens 0.41
GPT-4o May 2024 4 similar examples, strings 0.41
Llama 3.1-405B 4 random examples, strings 0.41
GPT-4o May 2024 4 similar examples, tokens 0.40
Llama 3.1-405B 4 random examples, tokens 0.40
Llama 3.1-70B 3 similar examples, strings 0.40
Llama 3.1-70B 4 similar examples, strings 0.40
Llama 3.1-70B 5 similar examples, strings 0.40
Llama 3.1-405B 3 random examples, strings 0.39
Llama 3.1-405B 5 random examples, tokens 0.39
Llama 3.1-405B 5 similar examples, tokens 0.39
Llama 3.1-70B 5 random examples, strings 0.39
GPT-4o May 2024 3 random examples, tokens 0.38
GPT-4o May 2024 3 similar examples, tokens 0.38
GPT-4o May 2024 3 similar examples, strings 0.38
GPT-4o mini 5 similar examples, strings 0.38
Llama 3.1-405B 3 similar examples, tokens 0.38
Llama 3.1-70B 4 random examples, tokens 0.37
Llama 3.1-70B 4 random examples, strings 0.37
GPT-4o mini 5 random examples, tokens 0.36
GPT-4o mini 4 similar examples, tokens 0.36
GPT-4o mini 5 similar examples, tokens 0.36
Llama 3.1-70B 3 random examples, strings 0.36
Llama 3.1-70B 5 random examples, tokens 0.36
Llama 3.1-70B 3 similar examples, tokens 0.36
Llama 3.1-70B 3 random examples, tokens 0.35
GPT-4o mini 4 random examples, tokens 0.33
GPT-4o mini 3 similar examples, tokens 0.33
GPT-4o mini 3 similar examples, strings 0.33
GPT-4o May 2024 5 random examples, strings 0.32
GPT-4o mini 3 random examples, tokens 0.32
GPT-4o mini 5 random examples, strings 0.32
GPT-4o mini 4 similar examples, strings 0.32
GPT-4o May 2024 4 random examples, strings 0.30
Llama 3.1-70B 5 similar examples, tokens 0.30
GPT-4o May 2024 3 random examples, strings 0.29
GPT-4o mini 4 random examples, strings 0.29
GPT-4o mini 3 random examples, strings 0.28
Llama 3.1-70B 4 similar examples, tokens 0.28

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 Jul 2024
GPT-4o 0.44

Source: The Climate-Change-NER 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
0.25 0.30 0.35 0.40 0.45 2024 0.25 0.30 0.35 0.40 0.45 2024

Source: Official Climate-Change-NER 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 Science and reasoning

All 203 in Science and reasoning
  • GPQA Diamond In use 198 graduate-level 'Google-proof' multiple-choice questions in biology, chemistry and physics. Gemini 4 Carbon 97%
  • Terminal-Bench-Science In use 70 scientific research-workflow tasks across five science areas, run in Harbor containers. Hidden tests give the resolution rate. GPT-6 Astra 68.1%
  • BioMysteryBench In use Bioinformatics problems on anonymized real datasets that require real analysis with tools. Rubric-graded, with human-solvable and human-difficult splits. Claude Fable 5.2 83.7%
  • FrontierScience In use Models solve physics, chemistry, and biology problems in an Olympiad track graded by short answer and a Research track graded by 10-point rubrics. Intern-S2-397B 87%
  • HiPhO In use Models solve theory problems from 2024 and 2025 physics olympiad exams, some with diagrams. Official marking schemes grade answers and steps, mapped to medal levels. Gemini 3 Pro 78.3%
  • LABBench2 In use 11 biology research sub-tasks (literature, figures, tables, patents, trials and more) with tool access. Macro-averaged accuracy. GPT-5.2 60.9%

Data audit

7 of 8 checks pass, 1 warning. Look first at the Newest score check. The newest came 2 years ago, in July 2024, for Llama 3.1-405B. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 1 warning 1 pass 3 notes
  • Pass Independent runs Scores from the benchmark’s paper (4 entries).
  • Note Models scored Only 4 models we track.
  • Warning Newest score The newest came 2 years ago, in July 2024, for Llama 3.1-405B.
  • Note Frontier models None of the top 10 has a score. The strongest tested is GPT-4o (#162).
  • Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on aclanthology.org: 48 rows for models we track at the last read.
  • 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