Gemini 2.5 Flash vs Gemini 2.5 Pro
Lado a lado. Cada métrica. Cada benchmark.
| Tipo | Gemini 2.5 Flash | Gemini 2.5 Pro |
|---|---|---|
| Provider |  Google DeepMind | Google DeepMind |
| pontuação média | 40.0 | 56.2 |
| Preço de entrada | $0.30 | $1.25 |
| Preço de saída | $2.50 | $10.00 |
| Janela de contexto | 1.0M tokens (~524 books) | 1.0M tokens (~524 books) |
| Lançado em | 2025-06-17 | 2025-06-17 |
| Código aberto | Proprietary | Proprietary |
Pontuações de benchmark
25 benchmarks · Gemini 2.5 Flash: 2, Gemini 2.5 Pro: 22
| Benchmark | Categoria | Gemini 2.5 Flash | Gemini 2.5 Pro |
|---|---|---|---|
| Aider polyglot | coding | 47.1 | 83.1 |
| ARC-AGI | reasoning | 32.3 | 41.0 |
| ARC-AGI-2 | reasoning | 2.5 | 4.9 |
| AudioMultiChallenge | knowledge | 40.0 | 46.9 |
| AudioMultiChallenge — Text Output | knowledge | 40.0 | 46.9 |
| Balrog | knowledge | 33.5 | 43.3 |
| Chatbot Arena Elo — Overall | arena | 1411.0 | 1448.2 |
| DeepResearch Bench | knowledge | 29.2 | 49.7 |
| Fiction.LiveBench | knowledge | 47.2 | 91.7 |
| FrontierMath-2025-02-28-Private | math | 4.8 | 14.1 |
| FrontierMath-Tier-4-2025-07-01-Private | math | 4.2 | 4.2 |
| GeoBench | knowledge | 73.0 | 81.0 |
| HELM — GPQA | knowledge | 39.0 | 74.9 |
| HELM — IFEval | language | 89.8 | 84.0 |
| HELM — MMLU-Pro | knowledge | 63.9 | 86.3 |
| HELM — Omni-MATH | math | 38.4 | 41.6 |
| HELM — WildBench | reasoning | 81.7 | 85.7 |
| HLE | knowledge | 7.7 | 17.7 |
| Lech Mazur Writing | knowledge | 76.5 | 86.0 |
| OTIS Mock AIME 2024-2025 | math | 73.0 | 84.7 |
| SimpleBench | reasoning | 29.4 | 54.9 |
| Terminal Bench | coding | 17.1 | 32.6 |
| The Agent Company | agentic | 41.1 | 30.3 |
| VPCT | knowledge | 7.0 | 19.6 |
| WeirdML | coding | 41.0 | 54.0 |