Gemini 2.5 Flash Lite (Thinking) — benchmark results
Gemini 2.5 Flash Lite evaluated with thinking enabled. Provider: Google. Released 2025-09-25. Access: API.
Unified ELO 1590 ± 8, rank #502 of 1776 rated models, from 299 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| EuroEval Danish NLU - ScaLA DA | 74.06 | Linguistic acceptability Score (%) | 98.7 |
| EuroEval Norwegian Common Sense Reasoning | 88.7 | Common Sense Reasoning Average Score (%) | 97.3 |
| EuroEval Bulgarian NLU - ScaLA BG | 75.58 | Linguistic acceptability Score (%) | 97.2 |
| EuroEval Ukrainian NLU - ScaLA UK | 45.21 | Linguistic acceptability Score (%) | 97 |
| EuroEval Czech NLU - CS GEC | 57.87 | Linguistic acceptability Score (%) | 96.1 |
| EuroEval Greek Summarization - Greek Wikipedia | 32.59 | Score (%) | 95.9 |
| EuroEval Romanian NLU - ScaLA RO | 60.68 | Linguistic acceptability Score (%) | 95.6 |
| EuroEval Bulgarian NLU | 70.77 | NLU Average Score (%) | 95.3 |
| EuroEval Norwegian NLU - ScaLA NB | 71.27 | Linguistic acceptability Score (%) | 95.3 |
| EuroEval Danish NLU | 65.3 | NLU Average Score (%) | 94.7 |
| EuroEval Slovene NLU - ScaLA SL | 58.26 | Linguistic acceptability Score (%) | 94.6 |
| EuroEval Bulgarian Knowledge | 75 | Knowledge Average Score (%) | 94.3 |
Interactive version: theaggregate.ai/model?slug=gemini-2-5-flash-lite-thinking · How the rankings work · Data refreshed daily, snapshot 2026-07-22.