Gemini 3 Flash Preview
von Google DeepMind · Veroeffentlicht 2025-12-17
49.1
Durchschn. Score
$0.50/1M
Eingabepreis
$3.00/1M
Ausgabepreis
1.0M tokens (~524 books)
Kontextfenster
multimodal
Typ
Tested on 24 benchmarks with 49.1% average. Top scores: Chatbot Arena Elo — Overall (1473.9%), Chatbot Arena Elo — Coding (1436.4%), OTIS Mock AIME 2024-2025 (92.8%).
Benchmark-Ergebnisse
| Benchmark | Kategorie | Score | Bar |
|---|---|---|---|
| Chatbot Arena Elo — Overall | arena | 1473.9 | |
| Chatbot Arena Elo — Coding | arena | 1436.4 | |
| OTIS Mock AIME 2024-2025 | math | 92.8 | |
| GeoBench | knowledge | 88.0 | |
| GPQA diamond | knowledge | 77.6 | |
| SWE-Bench verified | coding | 75.4 | |
| SimpleQA Verified | knowledge | 67.4 | |
| Terminal Bench | coding | 64.3 | |
| WeirdML | coding | 61.6 | |
| VPCT | knowledge | 58.9 | |
| MCP Atlas | agentic | 57.4 | |
| SimpleBench | reasoning | 53.3 | |
| Artificial Analysis — Agentic Index | speed | 49.7 | |
| Balrog | knowledge | 48.1 | |
| Artificial Analysis — Quality Index | speed | 46.4 | |
| Artificial Analysis — Coding Index | speed | 42.6 | |
| Chess Puzzles | knowledge | 38.0 | |
| FrontierMath-2025-02-28-Private | math | 35.6 | |
| ARC-AGI-2 | reasoning | 33.6 | |
| APEX-Agents | agentic | 24.0 | |
| SciPredict | knowledge | 22.2 | |
| ARC-AGI | reasoning | 21.5 | |
| GSO-Bench | coding | 9.8 | |
| FrontierMath-Tier-4-2025-07-01-Private | math | 4.2 |