Cross-lingual Question Answering on ECLeKTic
60.5Target AccuracyGEMINI-3-FLASH
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| GEMINI-3-FLASH2026.06 | 60.5 | 8 | 84.6 | |
| GEMINI-3-FLASHComplexity=LOW2026.06 | 60.2 | 8.9 | 84 | |
| GEMINI-3-FLASHComplexity=MINIMAL2026.06 | 52.3 | 13.4 | 74.2 | |
| CLAUDE-OPUS-4.72026.06 | 45.2 | 10.3 | 73.5 | |
| GEMINI-2.5-FLASH2026.06 | 44.5 | 6.4 | 81.7 | |
| CLAUDE-SONNET-4.62026.06 | 44.1 | 8.7 | 74.7 | |
| GEMINI-2.5-FLASHThinking=OFF2026.06 | 29.4 | 13.3 | 60.4 | |
| GPT-OSS-120B2026.06 | 25.4 | 2.4 | 82.2 | |
| GEMMA-4-31B-IT2026.06 | 22.6 | 5.5 | 68.6 | |
| CLAUDE-HAIKU-4.52026.06 | 20.8 | 8.4 | 59.8 | |
| GEMMA-4-26B-A4B-IT2026.06 | 20.3 | 5.7 | 63.9 | |
| GEMMA-3-27B2026.06 | 19.8 | 6.5 | 63.8 | |
| GPT-OSS-20B2026.06 | 18 | 1.6 | 80.1 | |
| QWEN-3-30B-A3B2026.06 | 16.6 | 2.8 | 65.9 | |
| GEMMA-3-12B2026.06 | 15.2 | 2.1 | 64.1 | |
| QWEN-3-4B2026.06 | 10.8 | 2.9 | 56.5 | |
| GEMMA-4-E4B-IT2026.06 | 10.7 | 5 | 58.4 | |
| GEMMA-3-4B2026.06 | 9.6 | 1.9 | 56.4 | |
| GEMMA-4-E2B-IT2026.06 | 9 | 2.8 | 51.8 | |
| GEMMA-3-1B2026.06 | 4.6 | 0.8 | 48.2 |