MMLU-ProVerified
89.5%
MMLU-Pro (AA run)
Models / Google
Gemini 3 Pro Preview (low)
Gemini 3 · Released 2025-11-18
AA Intelligence Index 33.1. Listed on Artificial Analysis with API pricing $2/$12 per 1M tokens. Benchmark rows sync from the AA snapshot — see the ledger on this page.
Frontier model tracked on Artificial Analysis. Use the benchmark ledger below for sourced scores; we do not infer numbers beyond published rows.
#88 of 131 on AA Index · current snapshot
33 AA Index · 13 benchmark rows · 1 sourcesPosition is benchmark-specific — not a cross-family or cross-source ranking.Composite60
ContextUnlisted
Input / 1M$2.0
Output / 1M$12
Knowledge cutoffunpublished
Output speed0 t/s
TTFT (default API)0.00s
Statussolid
- Knowledge
- 90
- Reasoning
- 89
- Math
- 87
- Coding
- 86
- Agentic
- 68
- Long context
- 67
81average across 6 tested families
Benchmark placements
Per-benchmark positions
Where Gemini 3 Pro Preview (low) places on each public benchmark source that publishes a row. Rank counts every model with a latest row in the same test — not a universal quality score.
| Benchmark | Family | Rank | Score | Source | Date |
|---|---|---|---|---|---|
| MMLU-Pro | Knowledge | #3/ 34 | 90 | Artificial Analysis | 2026-07-28 |
| LiveCodeBench | Coding | #8/ 26 | 86 | Artificial Analysis | 2026-07-28 |
| AIME | Math | #22/ 32 | 87 | Artificial Analysis | 2026-07-28 |
| SciCode | Coding | #40/ 144 | 50 | Artificial Analysis | 2026-07-28 |
| GPQA Diamond | Reasoning | #45/ 149 | 89 | Artificial Analysis | 2026-07-28 |
| AA-LCR | Long context | #46/ 143 | 67 | Artificial Analysis | 2026-07-28 |
| τ³-Bench Banking | Agentic | #46/ 143 | 68 | Artificial Analysis | 2026-07-28 |
| Humanity's Last Exam | Reasoning | #63/ 144 | 28 | Artificial Analysis | 2026-07-28 |
| AA output speed | Performance | #93/ 149 | Artificial Analysis | 2026-07-28 | |
| AA time to first token | Performance | #93/ 149 | Artificial Analysis | 2026-07-28 | |
| IFBench | Reasoning | #93/ 121 | 50 | Artificial Analysis | 2026-07-28 |
| Artificial Analysis Intelligence Index | Reasoning | #102/ 147 | 33 | Artificial Analysis | 2026-07-28 |
| Terminal-Bench | Agentic | #114/ 148 | 34 | Artificial Analysis | 2026-07-28 |
Family context
Strongest rows in each family
The three highest-scoring carded models in each capability family. Where Gemini 3 Pro Preview (low) shows up, it's highlighted.
KnowledgeGemini 3 Pro Preview (low) · 90
- 191
- 291
- 391
ReasoningGemini 3 Pro Preview (low) · 89
- 195
- 294
- 394
MathGemini 3 Pro Preview (low) · 87
- 199
- 299
- 399
CodingGemini 3 Pro Preview (low) · 86
- 196
- 296
- 396
AgenticGemini 3 Pro Preview (low) · 68
- 1100
- 2100
- 3100
Multimodal
- 172
- 272
- 371
Long contextGemini 3 Pro Preview (low) · 67
- 176
- 276
- 375
Tool use
- 189
- 287
- 387
Safety
- 192
- 292
- 391
Human preference
- 195
- 292
- 391
Newest receipts
13 recentLast 120 days
Benchmark rows added to the public ledger for Gemini 3 Pro Preview (low) in the last 120 days. Older rows live in the full table below.
| Benchmark | Family | Score | Source | Days ago |
|---|---|---|---|---|
| AA-LCR | Long context | 67.33% | Artificial Analysis | 4 |
| AA output speed | Performance | 0 t/s | Artificial Analysis | 4 |
| AA time to first token | Performance | 0.00s | Artificial Analysis | 4 |
| AIME | Math | 86.67% | Artificial Analysis | 4 |
| Artificial Analysis Intelligence Index | Reasoning | 33.1% | Artificial Analysis | 4 |
| GPQA Diamond | Reasoning | 88.7% | Artificial Analysis | 4 |
| Humanity's Last Exam | Reasoning | 27.6% | Artificial Analysis | 4 |
| IFBench | Reasoning | 49.66% | Artificial Analysis | 4 |
| LiveCodeBench | Coding | 85.7% | Artificial Analysis | 4 |
| MMLU-Pro | Knowledge | 89.5% | Artificial Analysis | 4 |
| SciCode | Coding | 49.9% | Artificial Analysis | 4 |
| τ³-Bench Banking | Agentic | 68.13% | Artificial Analysis | 4 |
| Terminal-Bench | Agentic | 34.09% | Artificial Analysis | 4 |
Family coverage
Best score per family
A high composite that hides a weak family is a trap. These bars surface the families where this model hasn't been publicly tested, and where it leads.
Price vs. performance
Where Gemini 3 Pro Preview (low) sits on the scatter
Hover or tab any dot for name, score, and input price.
AnthropicOpenAIMistralAlibabaDeepSeekGoogleMoonshotxAIMiniMaxMetaZhipuXiaomiOtherMeituanCohereNVIDIA
Full benchmark ledger
Every catalog benchmark for Gemini 3 Pro Preview (low). Scores link to the original source; gaps mean no public row exists yet.
13 of 32 catalog benchmarks have a sourced row for Gemini 3 Pro Preview (low).
Knowledge
Reasoning
33.1%
AA Intelligence Index v4.1
GPQA DiamondVerified
88.7%
GPQA Diamond (AA run)
Humanity's Last ExamVerified
27.6%
HLE (AA run)
IFBenchVerified
49.66%
IFBench (AA run)
Math
AIMEVerified
86.67%
AIME 2025 (AA run)
Coding
LiveCodeBenchVerified
85.7%
LiveCodeBench (AA run)
SciCodeVerified
49.9%
SciCode (AA run)
Agentic
Terminal-BenchVerified
34.09%
Terminal-Bench Hard (AA run)
τ³-Bench BankingVerified
68.13%
τ²-Bench Banking (AA run; maps to τ³ explainer)
Long context
AA-LCRVerified
67.33%
AA-LCR (AA run)
Performance
AA output speedVerified
0 t/s
Median output tokens/s (1k prompt, default provider)
AA time to first tokenVerified
0.00s
Median time to first token (1k prompt, default provider)
Sources
Where the numbers come from
- Artificial Analysis models index· checked 2026-08-01