EuroEval Bulgarian NLU - ScaLA BG: leaderboard
The "ScaLA BG" subtask of EuroEval Bulgarian NLU. Bulgarian-language EuroEval natural-language-understanding suite, separating NLU task performance from the broader all-task EuroEval aggregate.
Metric: Matthews correlation coefficient (%). Source: euroeval.com. Saturation forecast: Estimated already saturated. 212 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Pro (Preview) | 83.73 | #66 |
| 2 | Gemini 3.1 Flash Lite (Preview) | 83.09 | #194 |
| 3 | Gemini 3 Flash (Preview) (Thinking) | 79.4 | #82 (Gemini 3 Flash (Preview)) |
| 4 | GPT-5.4 Mini (2026-03-17) (High) | 77.4 | #169 (GPT-5.4 Mini (2026-03-17)) |
| 5 | GPT-5.2 | 77.04 | #110 |
| 6 | Gemini 2.5 Flash (Thinking) | 76.7 | #245 (Gemini 2.5 Flash) |
| 7 | Gemini 2.5 Flash Lite (Thinking) | 75.58 | #422 (Gemini 2.5 Flash Lite) |
| 8 | Claude Sonnet 4.6 | 75.49 | #90 |
| 9 | Gemini 3 Flash (Preview) (Non-reasoning) | 74.59 | #82 (Gemini 3 Flash (Preview)) |
| 10 | GPT-4.1 | 73.45 | #248 |
| 11 | Gemini 2.5 Pro | 72.75 | #152 |
| 12 | Grok 4.20 Beta (0309) (Reasoning) | 70.6 | #204 (Grok 4.20 Beta (0309)) |
| 13 | Gemma 4 31B (IT) | 68.85 | #176 |
| 14 | GPT-5 Mini (2025-08-07) | 68.28 | #172 |
| 15 | Claude Sonnet 4.5 (Thinking) | 67.95 | #142 (Claude Sonnet 4.5) |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=euroeval-bulgarian-nlu-scala-bg · How It Works · Data refreshed daily, snapshot 2026-10-20.