Skip to content
aiqbench

Google · model passport

Gemini 3 Flash

gemini-3-flash
  • Cloud only
  • Proprietary licence
  • Released 2026-01-09(first seen on LMArena)
8/27axes covered
Overall · LMArena
1,473#38 of 396
Best percentile
90.7Overall
Cohort standing
67.4median 55.0

What the data says

Benchmarked on 8 of 27 tracked axes. Every figure below is the model's own published result on that board.

Strongest axes

  • 90.7 pct — Overall (1,473)
  • 90.4 pct — Hard prompts (1,493)
  • 88.6 pct — Instruction following (1,458)

Uneven: 69.6 percentile points between its best and worst axis — 90.7 pct on Overall versus 21.1 pct on Document Arena.

  • Overall90.7 pct1,473Verify →
  • Hard prompts90.4 pct1,493Verify →
  • Instruction following88.6 pct1,458Verify →
  • Coding84.9 pct1,506Verify →
  • Factuality Arena83.1 pct1,468Verify →
  • Vision Arena81.3 pct1,273Verify →
  • WebDev Arena51.8 pct1,440Verify →
  • Document Arena21.1 pct1,439Verify →
  • Search ArenaNot benchmarked
  • Agent ArenaNot benchmarked
  • GPQA DiamondNot benchmarked
  • AIME (math)Not benchmarked
  • FrontierMathNot benchmarked
  • SimpleQA VerifiedNot benchmarked
  • ARC-AGI-2Not benchmarked
  • Humanity's Last ExamNot benchmarked
  • SimpleBenchNot benchmarked
  • LiveBenchNot benchmarked
  • BFCL v4 (function calling)Not benchmarked
  • SWE-bench Verified (live board)Not benchmarked
  • Terminal-Bench 4.0Not benchmarked
  • MMLU-Pro (Artificial Analysis)Not benchmarked
  • LiveCodeBench (Artificial Analysis)Not benchmarked
  • IFBench (Artificial Analysis)Not benchmarked
  • HHEM Hallucination (Vectara)Not benchmarked
  • SWE-bench VerifiedNot benchmarked
  • Terminal-BenchNot benchmarked

Percentiles are the model's standing within each board's own population. Axes marked "not benchmarked" have no published result for this model — that is an absence of data, not a low score.

  • gemini-3-flash stands at 67.4 (raw mean 74) — measured on 8 of 27 axes.

    standingtemplateVerify →

  • gemini-3-flash stands at 67.4 versus a cohort median of 55 across 115 models released within 180 days — above the peer median.

    cohorttemplateVerify →

Every claim is a computed fact from the current board data, phrased by a low-temperature model call and rejected by a validation gate if any number is not in the source row. How the leaderboards work →