Skip to content
aiqbench

Google · model passport

Gemini 2 5 Flash Preview 09.2025

gemini-2-5-flash-preview-09-2025
  • Cloud only
  • Proprietary licence
  • Released 2025-09-01(from model name)
7/27axes covered
Overall · LMArena
1,403#144 of 396
Best percentile
66.4Instruction following
Cohort standing
50.2median 52.0

What the data says

Benchmarked on 7 of 27 tracked axes. Every figure below is the model's own published result on that board.

Weakest axes

  • 9.1 pct — Terminal-Bench (17.1%)

Uneven: 57.3 percentile points between its best and worst axis — 66.4 pct on Instruction following versus 9.1 pct on Terminal-Bench.

  • Instruction following66.4 pct1,400Verify →
  • Overall63.9 pct1,403Verify →
  • Hard prompts63.4 pct1,420Verify →
  • Vision Arena58 pct1,229Verify →
  • Coding56 pct1,427Verify →
  • Factuality Arena34.9 pct1,424Verify →
  • Terminal-Bench9.1 pct17.1%Verify →
  • WebDev ArenaNot benchmarked
  • Search ArenaNot benchmarked
  • Document 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

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-2.5-flash-preview-09-2025 stands at 50.2 (raw mean 50.2) — measured on 7 of 27 axes.

    standingtemplateVerify →

  • gemini-2.5-flash-preview-09-2025 stands at 50.2 versus a cohort median of 52 across 147 models released within 180 days — below 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 →