Benchmark Almanac

PoETa v2

GPT-4.1 76.2 leads since September 2025
Not in model cards

Models answer multiple-choice and open-ended questions in Portuguese. The hub port selects native Portuguese tasks covering inference, reading, exams, proverbs and toxicity.

On November 19, 2026, the repository listed GPT-4.1 first at 76.2, followed by GPT-4o at 75.2 and Sabia 3 at 72.2.

Made by
PoETa v2 team
Released
September 2025
Kind
Knowledge in many languages
Level
University
Scored
Checked automatically
Answers
Multiple choice
Tasks
All public
Measures
Capability
Language
Portuguese
In and out
Text → Text
Interaction
Single turn
Items from
Exams and contests
Run by
Anyone
Strongest model tested
GPT-4.1 #138 on the capability index
Last model added
April 2025
Scores refreshed
November 20th, 2026
27
Entries on its own leaderboard, 15 models we track

GPT-4.1 leads with 76.2 among models released by April 2025 GPT-4.1 leads with 76.2 among models released by April 2025

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. This view shows only models we track, each at its best.

The newest model with a score here came out in April 2025, 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-4.1 76.2
  2. 2 GPT-4o Mar 2025 75.2
  3. 3 Sabia 3 72.2
  4. 4 Qwen2.5-14B 71.0
  5. 5 Qwen3-14B 70.5
  6. 6 Qwen3-8B 64.7
  7. 7 Qwen2.5-7B 63.7
  8. 8 Falcon 3 10B 63.5
  9. 9 Qwen3-4B 59.2
  10. 10 Qwen2-7B 58.8
  11. 11 Falcon 3 7B 58.5
  12. 12 Llama 3.1-8B 53.5
  13. 13 Qwen2.5-3B 53.1
  14. 14 Qwen3-1.7B 45.1
  15. 15 Qwen2.5-1.5B 43.0
  16. 16 Llama 2-13B 41.5
  17. 17 Qwen 1 7B 41.1
  18. 18 Falcon 3 3B 38.6
  19. 19 Qwen2-1.5B 38.4
  20. 20 Curio 7B 34.8
  21. 21 Sabia 32.4
  22. 22 Llama 2-7B 29.5
  23. 23 Llama 1 7B 19.9
  24. 24 Qwen 1 1.8B 19.0
  25. 25 Falcon 3 1B 18.6
  26. 26 Curio 1.1B 14.7
  27. 27 Tinyllama 1T 10.2

Source: Official PoETa v2 leaderboard

#EntryScoreSettingDateSource
1 GPT-4.1 76.2 Apr 2025 Official leaderboard
2 GPT-4o Mar 2025 75.2 Mar 2025 Official leaderboard
3 Sabia 3 72.2 Official leaderboard
4 Qwen2.5-14B 71.0 Sep 2024 Official leaderboard
5 Qwen3-14B 70.5 Apr 2025 Official leaderboard
6 Qwen3-8B 64.7 Apr 2025 Official leaderboard
7 Qwen2.5-7B 63.7 Sep 2024 Official leaderboard
8 Falcon 3 10B 10B 63.5 Official leaderboard
9 Qwen3-4B 59.2 Apr 2025 Official leaderboard
10 Qwen2-7B 58.8 Jun 2024 Official leaderboard
11 Falcon 3 7B 7B 58.5 Official leaderboard
12 Llama 3.1-8B 53.5 Jul 2024 Official leaderboard
13 Qwen2.5-3B 53.1 Sep 2024 Official leaderboard
14 Qwen3-1.7B 45.1 Apr 2025 Official leaderboard
15 Qwen2.5-1.5B 43.0 Sep 2024 Official leaderboard
16 Llama 2-13B 41.5 Jul 2023 Official leaderboard
17 Qwen 1 7B 7B 41.1 Official leaderboard
18 Falcon 3 3B 3B 38.6 Official leaderboard
19 Qwen2-1.5B 38.4 Jun 2024 Official leaderboard
20 Curio 7B 7B 34.8 Official leaderboard
21 Sabia 7B 32.4 Official leaderboard
22 Llama 2-7B 29.5 Jul 2023 Official leaderboard
23 Llama 1 7B 7B 19.9 Official leaderboard
24 Qwen 1 1.8B 1.8B 19.0 Official leaderboard
25 Falcon 3 1B 1B 18.6 Official leaderboard
26 Curio 1.1B 1.1B 14.7 Official leaderboard
27 Tinyllama 1T 1.1B 10.2 Official leaderboard
#ModelScoreSettingDateSource
1 GPT-4.1 76.2 Apr 2025 Official leaderboard
2 GPT-4o Mar 2025 75.2 Mar 2025 Official leaderboard
3 Qwen2.5-14B 71.0 Sep 2024 Official leaderboard
4 Qwen3-14B 70.5 Apr 2025 Official leaderboard
5 Qwen3-8B 64.7 Apr 2025 Official leaderboard
6 Qwen2.5-7B 63.7 Sep 2024 Official leaderboard
7 Qwen3-4B 59.2 Apr 2025 Official leaderboard
8 Qwen2-7B 58.8 Jun 2024 Official leaderboard
9 Llama 3.1-8B 53.5 Jul 2024 Official leaderboard
10 Qwen2.5-3B 53.1 Sep 2024 Official leaderboard
11 Qwen3-1.7B 45.1 Apr 2025 Official leaderboard
12 Qwen2.5-1.5B 43.0 Sep 2024 Official leaderboard
13 Llama 2-13B 41.5 Jul 2023 Official leaderboard
14 Qwen2-1.5B 38.4 Jun 2024 Official leaderboard
15 Llama 2-7B 29.5 Jul 2023 Official leaderboard

Number one over time

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

OpenAI
Own leaderboard Not tested since Sep 2025
GPT-4.1 76.2

Source: Official PoETa v2 leaderboard

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
On the benchmark’s own leaderboard
20.0 40.0 60.0 80.0 2024 2025 20.0 40.0 60.0 80.0 2024 2025

Source: Official PoETa v2 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 Multilingual

All 545 in Multilingual
  • MMMLU In use MMLU professionally translated into 14 languages. Multiple-choice accuracy. Claude Mythos Preview 92.7%
  • WMT24++ In use English-to-55-language translation of WMT24 test data. Scored with automatic metrics such as MetricX or xCOMET. K-EXAONE 90.5%
  • MMLU-ProX In use MMLU-Pro translated into up to 29 languages with expert review. Multiple-choice accuracy per language. Qwen3.7-Max 87%
  • CLIcK In use Models answer 1,995 multiple-choice questions on Korean language and culture. Questions come from official Korean exams and textbooks in eleven categories. Solar Open 2 90.7%
  • Global MMLU Lite 3.0 In use Models answer 400 human-translated MMLU questions per language in 23 languages. Version 3.0 adds Czech, Hungarian, Oriya, Slovak and Tajik and updates Italian. Accuracy. Claude Opus 5 95.7%
  • MMMB In use Models answer multiple-choice questions about images in six languages across 15 categories. The set has about 12,000 questions. LFM2.5-VL-3B 83%

Data audit

8 of 9 checks pass, 1 warning. Look first at the Newest score check. The newest came 19 months ago, in April 2025, for Qwen3-1.7B. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 1 warning 2 pass 2 notes
  • Pass Independent runs Scores from the official leaderboard (27 entries).
  • Pass Models scored 15 models we track, plus 12 other entries on its leaderboard.
  • Warning Newest score The newest came 19 months ago, in April 2025, for Qwen3-1.7B.
  • Note Frontier models None of the top 10 has a score. The strongest tested is GPT-4.1 (#138).
  • Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on github.com: 15 rows for models we track and 12 other entries 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