Benchmark Almanac

IberoBench

Llama 3-8B 30.3 leads since January 2025
Not in model cards

Models perform language understanding and generation tasks in Iberian languages. The suite combines existing evaluations with new tests specific to each language.

The linked harness folder contains the Basque tasks. The paper describes the complete suite.

Released
January 2025
Kind
Reasoning in many languages
Level
University
Scored
Checked automatically
Answers
Open-ended
Tasks
All public
Measures
Capability
Language
Many languages
In and out
Text → Text
Interaction
Single turn
Refresh
One fixed set
Run by
Anyone
In views
Many languages
Strongest model tested
Llama 3-8B #206 on the capability index
Last model added
April 2024
Scores refreshed
November 21st, 2026
33
Entries in its paper, 5 models we track

Llama 3-8B leads with 30.3 among models released by April 2024 Llama 3-8B leads with 30.3 among models released by April 2024

Best score per entry from the benchmark's paper, read on November 21st, 2026. A paper’s table doesn’t add newer models. The paper reports five language scores, so this is their plain mean. 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 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 Llama 3-8B 5-shot 30.3
  2. 2 Gemma 7B 5-shot 30.2
  3. 3 Mistral 7B v0.1 5-shot 27.9
  4. 4 Occiglot_eu5-7B 5-shot 26.5
  5. 5 Occiglot_es/en-7B 5-shot 25.7
  6. 6 Latxa-13B 5-shot 25.0
  7. 7 Tower-13B 5-shot 24.9
  8. 8 Tower-7B 5-shot 20.9
  9. 9 CataLlama-8B 5-shot 20.7
  10. 10 Latxa-7B 5-shot 20.1
  11. 11 OLMo-7B 5-shot 19.3
  12. 12 Gemma 2B 5-shot 18.7
  13. 13 Sabiá-7B 5-shot 17.9
  14. 14 BLOOM-7.1B 5-shot 17.1
  15. 15 FLOR-6.3B 5-shot 16.9
  16. 16 Aitana-6.3B 5-shot 15.5
  17. 17 BLOOM-3B 5-shot 15.1
  18. 18 Falcon-7B 5-shot 14.8
  19. 19 BLOOM-1.7B 5-shot 13.9
  20. 20 FLOR-1.3B 5-shot 12.1
  21. 21 XGLM-7.5B 5-shot 11.9
  22. 22 FLOR-760M 5-shot 11.6
  23. 23 BLOOM-1.1B 5-shot 11.6
  24. 24 XGLM-2.9B 5-shot 11.5
  25. 25 XGLM-4.5B 5-shot 10.6
  26. 26 Carballo-B-1.3B 5-shot 10.4
  27. 27 mGPT-13B 5-shot 9.54
  28. 28 mGPT-1.3B 5-shot 9.26
  29. 29 Carballo-C-1.3B 5-shot 7.80
  30. 30 Carvalho-1.3B 5-shot 7.62
  31. 31 XGLM-1.7B 5-shot 6.81
  32. 32 GlorIA-1.3B 5-shot 5.05
  33. 33 XGLM-564M 5-shot 4.80

Source: The IberoBench paper

#EntryScoreSettingDateSource
1 Llama 3-8B 30.3 5-shot Apr 2024 Official leaderboard
2 Gemma 7B 30.2 5-shot Feb 2024 Official leaderboard
3 Mistral 7B v0.1 27.9 5-shot Sep 2023 Official leaderboard
4 Occiglot_eu5-7B 26.5 5-shot Official leaderboard
5 Occiglot_es/en-7B 25.7 5-shot Official leaderboard
6 Latxa-13B 25.0 5-shot Official leaderboard
7 Tower-13B 24.9 5-shot Official leaderboard
8 Tower-7B 20.9 5-shot Official leaderboard
9 CataLlama-8B 20.7 5-shot Official leaderboard
10 Latxa-7B 20.1 5-shot Official leaderboard
11 OLMo-7B 19.3 5-shot Official leaderboard
12 Gemma 2B 18.7 5-shot Feb 2024 Official leaderboard
13 Sabiá-7B 17.9 5-shot Official leaderboard
14 BLOOM-7.1B 17.1 5-shot Official leaderboard
15 FLOR-6.3B 16.9 5-shot Official leaderboard
16 Aitana-6.3B 15.5 5-shot Official leaderboard
17 BLOOM-3B 15.1 5-shot Official leaderboard
18 Falcon-7B 14.8 5-shot Apr 2023 Official leaderboard
19 BLOOM-1.7B 13.9 5-shot Official leaderboard
20 FLOR-1.3B 12.1 5-shot Official leaderboard
21 XGLM-7.5B 11.9 5-shot Official leaderboard
22 FLOR-760M 11.6 5-shot Official leaderboard
23 BLOOM-1.1B 11.6 5-shot Official leaderboard
24 XGLM-2.9B 11.5 5-shot Official leaderboard
25 XGLM-4.5B 10.6 5-shot Official leaderboard
26 Carballo-B-1.3B 10.4 5-shot Official leaderboard
27 mGPT-13B 9.54 5-shot Official leaderboard
28 mGPT-1.3B 9.26 5-shot Official leaderboard
29 Carballo-C-1.3B 7.80 5-shot Official leaderboard
30 Carvalho-1.3B 7.62 5-shot Official leaderboard
31 XGLM-1.7B 6.81 5-shot Official leaderboard
32 GlorIA-1.3B 5.05 5-shot Official leaderboard
33 XGLM-564M 4.80 5-shot Official leaderboard
#ModelScoreSettingDateSource
1 Llama 3-8B 30.3 5-shot Apr 2024 Official leaderboard
2 Gemma 7B 30.2 5-shot Feb 2024 Official leaderboard
3 Mistral 7B v0.1 27.9 5-shot Sep 2023 Official leaderboard
4 Gemma 2B 18.7 5-shot Feb 2024 Official leaderboard
5 Falcon-7B 14.8 5-shot Apr 2023 Official leaderboard

Rankings by part

We read 5 parts from the benchmark’s paper, each ranking models on one part of the test, such as Catalan, European Spanish and Basque. Llama 3-8B leads 3 of them, and 2 other models lead the rest. Arrows show how far a model moved from its overall rank among the models we track.

Catalan

5 models we track have a score here.

  1. 1 Mistral 7B v0.1 ↑2 38.7
  2. 2 Llama 3-8B ↓1 37.9
  3. 3 Gemma 7B ↓1 36.1
  4. 4 Gemma 2B 22.1
  5. 5 Falcon-7B 17.9

Number one over time

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

Meta
Benchmark paper Not tested since Jan 2025
Llama 3-8B 30.3

Source: The IberoBench 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
10.0 15.0 20.0 25.0 30.0 35.0 2023 2024 10.0 15.0 20.0 25.0 30.0 35.0 2023 2024

Source: Official IberoBench 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 2 years ago, in April 2024, for Llama 3-8B. 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 benchmark’s paper (33 entries).
  • Pass Models scored 5 models we track, plus 28 other entries on its leaderboard.
  • Warning Newest score The newest came 2 years ago, in April 2024, for Llama 3-8B.
  • Note Frontier models None of the top 10 has a score. The strongest tested is Llama 3-8B (#206).
  • Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on aclanthology.org: 5 rows for models we track and 28 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