Home/Models/Gemini 3.5 Flash
Google DeepMind logo

Gemini 3.5 Flash

by Google DeepMind · Released May 2026

Multimodal1M Context
61.3
avg score
Rank #79
Compare
Better than 75% of all models
Context
1.0M tokens (~524 books)
Input $/1M
$1.50
Output $/1M
$9.00
Type
multimodal
License
Proprietary
Benchmarks
26 tested
Data as of
About

Gemini 3.5 Flash is Google's high-efficiency multimodal model, bringing near-Pro level coding and reasoning at Flash-tier cost and speed. It is highly optimized for coding proficiency and parallel agentic execution...

Tested on 26 benchmarks · BenchGecko score 61.3. Top scores: Chatbot Arena Elo — Coding (1505.5%), Chatbot Arena Elo — Overall (1476.3%), OTIS Mock AIME 2024-2025 (95.5%).

Looking for similar performance at lower cost?
gpt-oss-20b (free) scores 61.0 (100% as good) at $0.00/1M input · 100% cheaper
Capabilities
coding
59.8
#48 globally
reasoning
78.9
#19 globally
math
47.7
#123 globally
knowledge
68.0
#26 globally
agentic
27.5
#40 globally
speed
52.6
#23 globally
general
44.3
#53 globally
Benchmark Scores
Compare All
Tested on 26 benchmarks · Ranked across 8 categories
Score Distribution (all 312 models)
0255075100
▲ You are here
SWE-Bench verified

Real-world software engineering tasks from GitHub issues. Models must diagnose bugs and write patches that pass test suites. Human-verified subset of SWE-bench.

79.3·
WeirdML

Unusual and adversarial machine learning challenges. Tests robustness of reasoning about edge cases in ML systems.

62.6·
37.4·
ARC-AGI

Abstraction and Reasoning Corpus. Tests fluid intelligence through novel visual pattern recognition puzzles. Core measure of general intelligence.

92.5·
ARC-AGI-2

ARC-AGI 2, harder sequel to ARC. More complex abstract reasoning patterns that test generalization ability beyond training data.

72.1·
SimpleBench

Deceptively simple questions that humans find easy but AI models often get wrong. Tests common sense and reasoning gaps.

72.0·
OTIS Mock AIME 2024-2025

Mock AIME (American Invitational Mathematics Exam) problems from OTIS. Tests mathematical competition performance.

95.5·
FrontierMath-2025-02-28-Private

Original research-level math problems created by professional mathematicians. Problems are unpublished and cannot be memorized.

39.0·
Excellent (85+) Good (70-85) Average (50-70) Below (<50)
Links
Documentation
BenchGecko API
gemini-3-5-flash
Specifications
  • Typemultimodal
  • Context1.0M tokens (~524 books)
  • ReleasedMay 2026
  • LicenseProprietary
  • StatusActive
  • Cost / Message~$0.012
Available On
Google DeepMind logoGoogle DeepMind$1.50
Share & Export
Tweet
Gemini 3.5 Flash is a proprietary multimodal AI model by Google DeepMind, released in May 2026. It has an average benchmark score of 61.3. Context window: 1M tokens.

Key facts · as of 2026-10-05

  • Gemini 3.5 Flash by Google DeepMind. BenchGecko score 61.3, rank 79 of 312 scored models (normalized average of public benchmark scores).
  • List price $1.50 input · $9.00 output per 1M tokens (as of 2026-10-05).
  • Gecko Tests: Who Are You B (Knows who made it).

How to cite · data as of 2026-10-05

Gemini 3.5 Flash · benchmarks, pricing and providers. BenchGecko, data as of 2026-10-05. https://benchgecko.ai/model/gemini-3-5-flash

Credit "Source: BenchGecko" with a link. Prices per provider and Gecko Tests are BenchGecko data (CC BY 4.0); benchmark scores keep their original source, listed in the JSON. JSON · llms.txt · MCP