Gemini 2.0 Flash Lite: benchmark results
Google's most cost-efficient Gemini 2.0 tier, keeping the 1M-token multimodal context at ultra-low per-token prices (February 2025). Provider: Google. Released 2025-02-25. Access: API.
Aggregate Rating 1598 ± 9, rank #450 of 1656 rated models, from 424 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| ATAD - Referential Ambiguity | 62.75 | Accuracy (%) on the T5 referential ambiguity task (pick the | 100 |
| HELM Audio - Mustard | 66.09 | EM | 100 |
| LLM Stats (Bird-SQL (dev)) | 57.4 | Score (%) | 100 |
| ATAD - Logical Contradiction | 62 | Accuracy (%) on the T6 logical contradiction task (pick the | 95.5 |
| HELM AFR - MMLU Clinical Knowledge (Bambara) | 60 | EM | 95.2 |
| HELM AFR - MMLU Clinical Knowledge (Igbo) | 63.4 | EM | 95.2 |
| HELM AFR - MMLU Clinical Knowledge (Tswana) | 70.19 | EM | 95.2 |
| HELM SeaHELM - TyDiQA | 65.01 | SQuAD macro-averaged F1 score | 95 |
| HELM VHELM - Blink Reasoning - IQ Test | 50.67 | Prefix Quasi-Exact Match (%) | 93.6 |
| HELM VHELM - Mmmu - Clinical Medicine | 76.67 | Prefix Quasi-Exact Match (%) | 93.6 |
| HELM VHELM - Mmmu - Geography | 74.07 | Prefix Quasi-Exact Match (%) | 93.6 |
| MERA - ruCodeEval | 66.77 | pass@1 (%) | 93.3 |
Interactive version: theaggregate.ai/model?slug=gemini-2-0-flash-lite · How It Works · Data refreshed daily, snapshot 2026-10-20.