GoogleUnited States· released Dec 2025

Gemini 3 Flash

Name: Gemini 3 Flash
Author: Google

Ranked #26, 37.5% behind the frontier on average.

Google, United States. One of 14 models from this lab on the Index.

Third-party publishedMedium confidence (7+ comparable tests, 3+ Index categories)

Snapshot September 8, 2026

OpenCharts Index

62.5

Ranked #26 (21 to 30) of 36 · 37.5% behind the frontier on average · 18 comparable tests · 3 counting categories

Drop any single test and the Index lands between 57.0 and 66.6, anywhere from #21 to #30. Neighbours inside that range are ties.

5 of 8 hardest tests · share of the frontier

Distance to today's frontier on the hardest tests. Not a measure of intelligence.

What it's known for

Gemini 3 Flash, in 18 results.

Strongest in reasoning (71.8) and agentic (70.1). It trails the frontier most in math, 54.3% behind on average. No comparable result yet in long context. Between 35% and 40% behind the best published results, on average. Drop any single test and it would sit anywhere from #21 to #30.

Strongest category

Reasoning

Hard, novel problems: graduate science, abstract puzzles, expert exams.

71.8

GPQA Diamond89.4% · 91% of best
ARC-AGI-184.7% · 86% of best
SimpleBench61.1% · 75% of best
ARC-AGI-233.6% · 35% of best

Mean of 4 of the category's 6 comparable tests. Counts toward the Index.

Best single result

Mock AIME 2024–2025

Solve competition mathematics problems that have a single exact integer answer.

95.6%

95.6% of the best (100.0%) · #19 of 78.

Olympiad-style problems need long, careful chains of reasoning, and there is no partial credit.

Where it trails

ProofBench

Write a complete, rigorous proof, graded for rigor rather than for a final answer.

15.0%

85.0% behind the best published result (100.0%) · #44 of 61.

A proof shows the reasoning itself. A right answer with a wrong argument scores nothing.

Not measured yet

No comparable result yet in this category. A blank is a blank, never a zero, and it does not lower the Index.

Long context

Covered, but not averaged into the Index:

Knowledgesingle-test category, shown beside the Index · 88.4
Coding3 of 7 tests · needs 4 · mean 59.7
Multimodalsingle-test category, shown beside the Index · 88.4
Human preferencesingle-test category, shown beside the Index · 90.4

Every score, with its source

See it at work

What Gemini 3 Flash was asked to do, and how it did.

Every test it has taken, by category, with the kind of task it faced, who produced the number and where the result landed against the best published one. The examples are original and illustrative, never items from the datasets themselves.

Human preference · mean 90.4 across 1 comparable test · a single-test category, shown beside the Index and never averaged into it

Text ArenaLeaderboard

Answer the same prompt as a rival model; real users vote blind on which answer they prefer.

1,474

45% expected win rate against the board leader (1,507)
#24 of 92

board 2026-09-02

Source: Arena

Every result is read against the best published one on its test: percent scores as a share of the best above chance, Arena ratings as the expected win rate against the board leader, open-ended values on a log scale. Sitting a harder exam never lowers a model. Every number links to the publisher that produced it.

Head to head

Gemini 3 Flash against whoever you pick.

Choose a rival. Every counted test both have taken appears side by side as a share of the best published result on that test, with the category means above. A win is a higher share; a gap under 2 points is a tie, because one item on a 45-problem exam is 2.2 points and publishers' own standard errors are of that order.

Gemini 3 Flash

#26·Index 62.5C

0 : 11

wins · 0 ties · 11 shared

GPT-6 Astra

#1·Index 99.0A+

GPT-6 Astra comes out ahead on 11 of the 11 tests both have taken.

Category means

Reasoning
71.8
99.6
Knowledge
88.4
100.0
Coding
59.7
97.7
Math
45.7
99.8
Agentic
70.1
98.5
Multimodal
88.4
—
Human preference
90.4
—

Shared tests · biggest gaps first

ProofBenchMath
15.0%
99.0%
−84.0
FrontierMath Tier 4Math
17.1%
97.6%
−82.5
Code Arena (WebDev)Coding
1,438
1,797
−77.6
ARC-AGI-2Reasoning
33.6%
95.0%
−64.6
APEX-AgentsAgentic
24.0%
46.7%
−47.9
FrontierMath (Tiers 1–3)Math
51.2%
93.7%
−45.3
WeirdMLCoding
61.6%
92.9%
−33.7
ARC-AGI-1Reasoning
84.7%
98.5%
−14.0
SimpleQA VerifiedKnowledge
66.8%
75.6%
−11.6
GPQA DiamondReasoning
89.4%
95.8%
−9.0
Mock AIME 2024–2025Math
95.6%
100.0%
−4.4

Bars are distances to the best published result on each test (percent scores above chance, Arena ratings as win rate against the leader, open-ended values on a log scale). Raw values as the publishers report them.

Open GPT-6 Astra

Scores

Every test, every source

The best published run per benchmark, the raw value as the publisher reports it, who produced the number, how close it comes to the best published result on that test, and where the model sits among every model scored on it. Percent scores are read above chance, Arena ratings as a win rate against the board leader, open-ended values on a log scale. A test with too few models to compare against is shown but not counted.

Category means

Reasoning71.8 · 4/6
Knowledgebeside88.4 · 1/1
Coding59.7 · 3/7
Math45.7 · 4/5
Agentic70.1 · 4/8
Multimodalbeside88.4 · 1/1
Human preferencebeside90.4 · 1/1
Long context—

Mean · tests taken of the category's comparable tests. A faded category either rests on fewer than half of them or is a single-test category (marked “beside”), which is shown next to the Index and never averaged into it.

GPQA DiamondEpoch-runReasoning

89.4%

91.0% of the best (95.8%), above chance
58th percentile of 78 · n = 198

preview · high

Gemini 3 Flash

Gemini 3 Flash, in 18 results.

What Gemini 3 Flash was asked to do, and how it did.

Gemini 3 Flash against whoever you pick.

Every test, every source

Compare with its neighbours.

The best models, ranked here, working inside Theo.