LA Leaderboard - QNLI Basque: leaderboard

Question-sentence entailment in Basque: whether a sentence answers a question (QNLIeu). From the LA Leaderboard, which evaluates models on Spanish, Catalan, Basque and Galician tasks with the lm-evaluation-harness; accuracy (%), higher is better. Each model result file is a separate run.

Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 69 models tracked.

Top models

#ModelScoreOverall rank
1Gemma 2 9B (IT)66.81#794
2Gemma 3 4B (IT)59.66#996
3gemma-3-4B-pt Base57.98#1110
4Qwen 3 8B57.14#679
5Gemma 2 9B57.14#924
6Qwen3-8B-Base56.3#627
7salamandra-7B55.46#1447
8Teuken-7B-instruct-commercial-v0.455.46#1422
9aya-expanse-8B55.04#915
10Mistral 7B Instruct (v0.3)54.62#1256
11SmolLM2-1.7B-Instruct54.62#1542
12Llama 3.1 8B Instruct54.2#1019
13Gemma 3 1B (IT)54.2#1482
14DeepSeek R1 Distill Qwen 1.5B53.78#1522
15SmolLM3-3B53.78#1238

No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #6 GPT-6.1 Sol.

Interactive version: theaggregate.ai/benchmark?slug=la-leaderboard-qnli-basque · How It Works · Data refreshed daily, snapshot 2026-11-06.