P3B3
Models hold conversations in Portuguese under three language instructions. The judge measures how closely replies follow European or Brazilian usage.
The workshop paper compares 20 models on 74 conversations with 203 turns. Higher scores indicate closer European usage under the European Portuguese instruction.
- Made by
- NOVA University of Lisbon
- Released
- March 2026
- Kind
- Chat and instructions in many languages
- Level
- Everyday
- Scored
- Graded by a model
- Answers
- Open-ended
- Tasks
- All public
- Measures
- Capability
- Subject
- Languages and literature
- Language
- Portuguese
- Interaction
- Multi-turn
- Items from
- Written by experts
- Refresh
- One fixed set
- Run by
- Anyone
- Strongest model tested
-
Gemini 3 Flash #60 on the capability index
- Last model added
- April 2026
- Scores refreshed
- November 20th, 2026
Gemini 3 Flash leads with 99.8 among models released by April 2026 Gemini 3 Flash leads with 99.8 among models released by April 2026
Best score per entry 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. 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 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- GPT-6.2 Astra
- Claude Fable 5.2
-
Gemini 4 Carbon
- GPT-6 Astra
- and 16 more
- 99.8
Gemini 3 Flash
- 98.0
Gemma 4 31B IT
- 95.9 AMALIA-9B
- 94.4
Gemma 4 E4B
- 93.3
Qwen3.5-27B
- 92.9
Gemma 3 12B
- 90.7
Ministral 3 14B
- 88.5 EuroLLM-22B
- 87.2 Sabiá-4
- 86.3
Apertus 70B
- 86.1
Qwen3.5-9B
- 75.2 Gervásio-70B-ptpt
- 67.8
Llama 3.3 70B
- 61.0
Apertus 8B Instruct
- 52.8 Salamandra-7B
- 47.7
Llama 3.1-8B
- 46.6 Gervásio-8B-ptpt
- 42.8
Olmo 3.1 32B Instruct
- 32.7
Qwen3-8B
- 23.3 OLMo-3-7B
- 1
Gemini 3 Flash pt-PT 99.8
- 2
Gemma 4 31B IT pt-PT 98.0
- 3 AMALIA-9B pt-PT 95.9
- 4
Gemma 4 E4B pt-PT 94.4
- 5
Qwen3.5-27B pt-PT 93.3
- 6
Gemma 3 12B pt-PT 92.9
- 7
Ministral 3 14B pt-PT 90.7
- 8 EuroLLM-22B pt-PT 88.5
- 9 Sabiá-4 pt-PT 87.2
- 10
Apertus 70B pt-PT 86.3
- 11
Qwen3.5-9B pt-PT 86.1
- 12 Gervásio-70B-ptpt pt-PT 75.2
- 13
Llama 3.3 70B pt-PT 67.8
- 14
Apertus 8B Instruct pt-PT 61.0
- 15 Salamandra-7B pt-PT 52.8
- 16
Llama 3.1-8B pt-PT 47.7
- 17 Gervásio-8B-ptpt pt-PT 46.6
- 18
Olmo 3.1 32B Instruct pt-PT 42.8
- 19
Qwen3-8B pt-PT 32.7
- 20 OLMo-3-7B pt-PT 23.3
Source: The P3B3 paper
- 1
Gemini 3 Flash pt-PT 99.8
- 2
Gemma 4 31B IT pt-PT 98.0
- 3
Gemma 4 E4B pt-PT 94.4
- 4
Qwen3.5-27B pt-PT 93.3
- 5
Gemma 3 12B pt-PT 92.9
- 6
Ministral 3 14B pt-PT 90.7
- 7
Apertus 70B pt-PT 86.3
- 8
Qwen3.5-9B pt-PT 86.1
- 9
Llama 3.3 70B pt-PT 67.8
- 10
Apertus 8B Instruct pt-PT 61.0
- 11
Llama 3.1-8B pt-PT 47.7
- 12
Olmo 3.1 32B Instruct pt-PT 42.8
- 13
Qwen3-8B pt-PT 32.7
Source: The P3B3 paper
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | | 99.8 | pt-PT | Dec 2025 | Official leaderboard |
| 2 | | 98.0 | pt-PT | Apr 2026 | Official leaderboard |
| 3 | AMALIA-9B open weights | 95.9 | pt-PT | Official leaderboard | |
| 4 | | 94.4 | pt-PT | Mar 2026 | Official leaderboard |
| 5 | | 93.3 | pt-PT | Feb 2026 | Official leaderboard |
| 6 | | 92.9 | pt-PT | Mar 2025 | Official leaderboard |
| 7 | | 90.7 | pt-PT | Dec 2025 | Official leaderboard |
| 8 | EuroLLM-22B open weights | 88.5 | pt-PT | Official leaderboard | |
| 9 | Sabiá-4 closed weights | 87.2 | pt-PT | Official leaderboard | |
| 10 | | 86.3 | pt-PT | Sep 2025 | Official leaderboard |
| 11 | | 86.1 | pt-PT | Feb 2026 | Official leaderboard |
| 12 | Gervásio-70B-ptpt open weights | 75.2 | pt-PT | Official leaderboard | |
| 13 | | 67.8 | pt-PT | Dec 2024 | Official leaderboard |
| 14 | | 61.0 | pt-PT | Sep 2025 | Official leaderboard |
| 15 | Salamandra-7B open weights | 52.8 | pt-PT | Official leaderboard | |
| 16 | | 47.7 | pt-PT | Jul 2024 | Official leaderboard |
| 17 | Gervásio-8B-ptpt open weights | 46.6 | pt-PT | Official leaderboard | |
| 18 | | 42.8 | pt-PT | Dec 2025 | Official leaderboard |
| 19 | | 32.7 | pt-PT | Apr 2025 | Official leaderboard |
| 20 | OLMo-3-7B open weights | 23.3 | pt-PT | Official leaderboard |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | | 99.8 | pt-PT | Dec 2025 | Official leaderboard |
| 2 | | 98.0 | pt-PT | Apr 2026 | Official leaderboard |
| 3 | | 94.4 | pt-PT | Mar 2026 | Official leaderboard |
| 4 | | 93.3 | pt-PT | Feb 2026 | Official leaderboard |
| 5 | | 92.9 | pt-PT | Mar 2025 | Official leaderboard |
| 6 | | 90.7 | pt-PT | Dec 2025 | Official leaderboard |
| 7 | | 86.3 | pt-PT | Sep 2025 | Official leaderboard |
| 8 | | 86.1 | pt-PT | Feb 2026 | Official leaderboard |
| 9 | | 67.8 | pt-PT | Dec 2024 | Official leaderboard |
| 10 | | 61.0 | pt-PT | Sep 2025 | Official leaderboard |
| 11 | | 47.7 | pt-PT | Jul 2024 | Official leaderboard |
| 12 | | 42.8 | pt-PT | Dec 2025 | Official leaderboard |
| 13 | | 32.7 | pt-PT | Apr 2025 | Official leaderboard |
Rankings by part
We read 2 parts from the benchmark’s paper: PeroVaz and PtBrVId. Each ranks models on one part of the test. Gemini 3 Flash leads both. Arrows show how far a model moved from its overall rank among the models we track.
PeroVaz
13 models we track have a score here.
- 1
Gemini 3 Flash 79.7
- 2
Qwen3.5-27B ↑2 78.4
- 3
Qwen3.5-9B ↑5 76.0
- 4
Ministral 3 14B ↑2 74.8
- 5
Gemma 3 12B 72.2
- 6
Gemma 4 E4B ↓3 70.1
- 7
Gemma 4 31B IT ↓5 69.7
- 8
Apertus 70B ↓1 68.8
- 9
Llama 3.3 70B 61.6
- 10
Apertus 8B Instruct 59.6
PtBrVId
13 models we track have a score here.
- 1
Gemini 3 Flash 99.3
- 2
Qwen3.5-27B ↑2 98.1
- 3
Gemma 4 E4B 96.4
- 4
Gemma 4 31B IT ↓2 95.6
- 5
Ministral 3 14B ↑1 95.2
- 6
Gemma 3 12B ↓1 95.0
- 7
Qwen3.5-9B ↑1 93.6
- 8
Apertus 70B ↓1 85.7
- 9
Llama 3.3 70B 81.5
- 10
Llama 3.1-8B ↑1 69.3
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.
Source: The P3B3 paper
| Model | Effort | Harness | Score | Cost per task | Output tokens |
|---|---|---|---|---|---|
| pt-PT | 99.8 | ||||
| pt-PT | 98.0 | ||||
| pt-PT | 94.4 | ||||
| pt-PT | 93.3 | ||||
| pt-PT | 92.9 | ||||
| pt-PT | 90.7 | ||||
| pt-PT | 86.3 | ||||
| pt-PT | 86.1 | ||||
| pt-PT | 67.8 | ||||
| No-Prompt | 63.7 | ||||
| pt-PT | 61.0 | ||||
| No-Prompt | 49.2 | ||||
| pt-PT | 47.7 | ||||
| No-Prompt | 44.5 | ||||
| pt-PT | 42.8 | ||||
| No-Prompt | 35.1 | ||||
| pt-PT | 32.7 | ||||
| No-Prompt | 23.4 | ||||
| No-Prompt | 21.9 | ||||
| No-Prompt | 20.3 | ||||
| No-Prompt | 17.8 | ||||
| No-Prompt | 16.6 | ||||
| No-Prompt | 16.1 | ||||
| No-Prompt | 15.3 | ||||
| pt-BR | 14.7 | ||||
| No-Prompt | 12.6 | ||||
| pt-BR | 12.2 | ||||
| pt-BR | 11.9 | ||||
| pt-BR | 10.6 | ||||
| pt-BR | 10.5 | ||||
| No-Prompt | 10.4 | ||||
| pt-BR | 8.60 | ||||
| pt-BR | 7.90 | ||||
| pt-BR | 7.30 | ||||
| pt-BR | 6.70 | ||||
| pt-BR | 4.90 | ||||
| pt-BR | 4.20 | ||||
| pt-BR | 4.20 | ||||
| pt-BR | 4.10 |
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.
Source: The P3B3 paper
| Model | Effort | Harness | Score | Cost per task | Output tokens |
|---|---|---|---|---|---|
| pt-PT | 99.8 | ||||
| pt-PT | 98.0 | ||||
| pt-PT | 94.4 | ||||
| pt-PT | 93.3 | ||||
| pt-PT | 92.9 | ||||
| pt-PT | 90.7 | ||||
| pt-PT | 86.3 | ||||
| pt-PT | 86.1 | ||||
| pt-PT | 67.8 | ||||
| No-Prompt | 63.7 | ||||
| pt-PT | 61.0 | ||||
| No-Prompt | 49.2 | ||||
| pt-PT | 47.7 | ||||
| No-Prompt | 44.5 | ||||
| pt-PT | 42.8 | ||||
| No-Prompt | 35.1 | ||||
| pt-PT | 32.7 | ||||
| No-Prompt | 23.4 | ||||
| No-Prompt | 21.9 | ||||
| No-Prompt | 20.3 | ||||
| No-Prompt | 17.8 | ||||
| No-Prompt | 16.6 | ||||
| No-Prompt | 16.1 | ||||
| No-Prompt | 15.3 | ||||
| pt-BR | 14.7 | ||||
| No-Prompt | 12.6 | ||||
| pt-BR | 12.2 | ||||
| pt-BR | 11.9 | ||||
| pt-BR | 10.6 | ||||
| pt-BR | 10.5 | ||||
| No-Prompt | 10.4 | ||||
| pt-BR | 8.60 | ||||
| pt-BR | 7.90 | ||||
| pt-BR | 7.30 | ||||
| pt-BR | 6.70 | ||||
| pt-BR | 4.90 | ||||
| pt-BR | 4.20 | ||||
| pt-BR | 4.20 | ||||
| pt-BR | 4.10 |
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: The P3B3 paper
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official P3B3 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
- 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 7 months ago, in April 2026, for Gemma 4 31B IT. 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 (20 entries).
- Pass Models scored 13 models we track, plus 7 other entries on its leaderboard.
- Warning Newest score The newest came 7 months ago, in April 2026, for Gemma 4 31B IT.
- Note Frontier models None of the top 10 has a score. The strongest tested is Gemini 3 Flash (#60).
- Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
- Pass Importer We read the board on arxiv.org: 39 rows for models we track and 7 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.