Gemini 3 Flash

GoogleReleased 17 Dec 2025ClosedSmall / fastDate verified

Compare Announcement

158147–169Index score#43 of 229 models

5/ 8Core tests taken4 of 6 areas · plus 16 anchor tests

−4Versus the frontier at releasebest then: GPT-5.2Replaced by Gemini 3.5 Flash

$0.50/ $3.0Price per million tokensinput / output, list price

1.05MContext windowtokens it reads at once

Where it sitson the frontier

Gemini 3 Pro reached this score 1 month earlier.

Test resultsby area · tick = best by any model

Reasoning & knowledgeevidence

GPQA Diamond

89%

Humanity's Last Exam

no result

Mathsevidence

FrontierMath T1-3

51%

Codingevidence

SWE-bench Verified

75%

Terminal-Bench

64%

Agentsevidence

APEX-Agents

no result

Novel problemsevidence

ARC-AGI-2

34%

Long tasksno evidence yet

METR time horizon

no result

Independent results from the Epoch AI benchmarking hub, best reasoning setting. The core tests are shown here; anchor tests also inform the score and can be shown above.

What others sayrank on each leaderboard

The Model Indexindependent tests, our method#43of 219 · 158

LMArena · Creative writingcreative writing#26of 386 · 1456 LMArena · Textoverall chat quality, judged by people#38of 388 · 1473 LMArena · Mathsmaths questions in chat#41of 373 · 1474 LMArena · Codingcoding questions in chat#60of 383 · 1505 Epoch Capabilities Indexoverall capability from many benchmarks#43of 239 · 151.8 LMArena · Visionunderstanding images#30of 149 · 1273 LMArena · WebDevbuilding web apps and front ends#64of 123 · 1439 LMArena · Documentsworking with long documents#38of 41 · 1413

Further left is better. Leaderboards measure different things: LMArena is people voting blind between two answers, the others are test scores. Click any row for the source.