25%
AA Intelligence Index v4.1
Modelle / Google
Gemini 3.1 · Release 2026-05-07
Googles günstigstes 1M-Kontext-Gemini. $0,25/$1,50 pro 1M Tokens. Vals-Index 35,24 (#16). Volumen-Tier für Klassifikation und Routing.
Hochvolumiges Preprocessing, Tagging und günstige RAG-Eingänge vor Eskalation zu Pro.
56average across 4 tested families
Wo Gemini 3.1 Flash Lite auf jeder öffentlichen Benchmark-Quelle mit Zeile steht. Rang zählt jedes Modell mit neuester Zeile im selben Test — kein universeller Qualitätsscore.
| Benchmark | Family | Rang | Score | Quelle | Datum |
|---|---|---|---|---|---|
| AA output speed | Performance | #4/ 69 | Artificial Analysis | 2026-07-21 | |
| IFBench | Reasoning | #8/ 57 | 77 | Artificial Analysis | 2026-07-21 |
| Vals Index | Agentic | #20/ 20 | 35 | Vals AI — Vals Index | 2026-06-04 |
| AA time to first token | Performance | #25/ 69 | Artificial Analysis | 2026-07-21 | |
| AA-LCR | Long context | #38/ 64 | 65 | Artificial Analysis | 2026-07-21 |
| GPQA Diamond | Reasoning | #44/ 69 | 82 | Artificial Analysis | 2026-07-21 |
| SciCode | Coding | #48/ 65 | 42 | Artificial Analysis | 2026-07-21 |
| Humanity's Last Exam | Reasoning | #49/ 65 | 16 | Artificial Analysis | 2026-07-21 |
| Artificial Analysis Intelligence Index | Reasoning | #52/ 67 | 25 | Artificial Analysis | 2026-07-21 |
| τ³-Bench Banking | Agentic | #58/ 64 | 9 | Artificial Analysis | 2026-07-21 |
| Terminal-Bench | Agentic | #59/ 69 | 31 | Artificial Analysis | 2026-07-21 |
Die drei höchstscorierenden kartierten Modelle pro Capability-Family. Wo Gemini 3.1 Flash Lite auftaucht, ist es hervorgehoben.
Benchmark-Zeilen im öffentlichen Ledger für Gemini 3.1 Flash Lite in den letzten 120 Tagen. Ältere Zeilen stehen in der vollen Tabelle unten.
| Benchmark | Family | Score | Quelle | Tage her |
|---|---|---|---|---|
| AA-LCR | Long context | 65.33% | Artificial Analysis | 2 |
| AA output speed | Performance | 315 t/s | Artificial Analysis | 2 |
| AA time to first token | Performance | 4.90s | Artificial Analysis | 2 |
| Artificial Analysis Intelligence Index | Reasoning | 25% | Artificial Analysis | 2 |
| GPQA Diamond | Reasoning | 82.2% | Artificial Analysis | 2 |
| Humanity's Last Exam | Reasoning | 16.2% | Artificial Analysis | 2 |
| IFBench | Reasoning | 77.21% | Artificial Analysis | 2 |
| SciCode | Coding | 41.9% | Artificial Analysis | 2 |
| τ³-Bench Banking | Agentic | 8.66% | Artificial Analysis | 2 |
| Terminal-Bench | Agentic | 31.09% | Artificial Analysis | 2 |
| Vals Index | Agentic | 35.24% | Vals AI — Vals Index | 49 |
Ein hoher Composite, der eine schwache Family versteckt, ist eine Falle. Diese Balken zeigen Families ohne öffentlichen Test — und wo das Modell führt.
Hover or tab any dot for name, score, and input price.
Jeder Katalog-Benchmark für Gemini 3.1 Flash Lite. Scores verlinken zur Originalquelle; Lücken heißen: noch keine öffentliche Zeile.
11 of 32 catalog benchmarks have a sourced row for Gemini 3.1 Flash Lite.
25%
AA Intelligence Index v4.1
82.2%
GPQA Diamond (AA run)
16.2%
HLE (AA run)
77.21%
IFBench (AA run)
41.9%
SciCode (AA run)
31.09%
Terminal-Bench v2.1 (AA run)
35.24%
Vals Index
8.66%
τ³-Bench Banking (AA run)
65.33%
AA-LCR (AA run)
315 t/s
Median output tokens/s (1k prompt, default provider)
4.90s
Median time to first token (1k prompt, default provider)