Benchfolio

Leaderboard / Model

Gemini 3.5 Flash

Google DeepMind · Closed weights · Released May 19, 2026 · $1.50 in · $9 out per 1M tokens

52Overall · rank 16

Score by job

Coding

55 · rank 19 · 3 of 5 benchmarks

BenchmarkResultVariantz
DeepSWE37.4%gemini-3.5-flash_medium-1.26
SWE-bench Verified79.3%gemini-3.5-flash_high0.90
LMArena Coding1491 ratinggemini-3.5-flash-high0.01

Agents

Not enough data · 1 of 4 benchmarks

BenchmarkResultVariantz
APEX-Agents27.5%gemini-3.5-flash_unknown-2.36

Reasoning

69 · rank 12 · 3 of 4 benchmarks

BenchmarkResultVariantz
GPQA Diamond92.8%gemini-3.5-flash_high0.50
ARC-AGI-272.1%gemini-3.5-flash_high0.37
SimpleBench76.7%gemini-3.5-flash_unknown1.13

Math

41 · rank 20 · 5 of 5 benchmarks

BenchmarkResultVariantz
FrontierMath Tiers 1–362.8%gemini-3.5-flash_high-0.36
FrontierMath Tier 426.8%gemini-3.5-flash_high-0.59
OTIS Mock AIME95.6%gemini-3.5-flash_high-0.00
ProofBench31.0%gemini-3.5-flash_high-0.66
LMArena Math1505 ratinggemini-3.5-flash-high0.92

Writing

68 · rank 12 · 2 of 2 benchmarks

BenchmarkResultVariantz
LMArena Creative Writing1469 ratinggemini-3.5-flash-high0.53
LMArena Instruction Following1466 ratinggemini-3.5-flash-high0.05

Design

28 · rank 35 · 1 of 1 benchmarks

BenchmarkResultVariantz
LMArena WebDev1500 ratinggemini-3.5-flash-high-0.57