NEWHow AI models draw the world map · Gecko Tests
Beta
Home/Models/Gemini 2.5 Flash
Google DeepMind logo

Gemini 2.5 Flash

by Google DeepMind · Released Jun 2025

Multimodal1M Context
37.3
avg score
Rank #211
Compare
Better than 32% of all models
Context
1.0M tokens (~524 books)
Input $/1M
$0.30
Output $/1M
$2.50
Type
multimodal
License
Proprietary
Benchmarks
28 tested
Data as of
About

Gemini 2.5 Flash is Google's state-of-the-art workhorse model, specifically designed for advanced reasoning, coding, mathematics, and scientific tasks. It includes built-in "thinking" capabilities, enabling it to provide responses with greater...

Tested on 28 benchmarks · BenchGecko score 37.3. Top scores: Chatbot Arena Elo — Overall (1409.4%), HELM — IFEval (89.8%), HELM — WildBench (81.7%).

Looking for similar performance at lower cost?
GPT-5 Nano scores 37.7 (101% as good) at $0.05/1M input · 83% cheaper
Capabilities
coding
35.0
#154 globally
reasoning
36.5
#111 globally
math
31.7
#181 globally
knowledge
41.5
#193 globally
agentic
21.4
#45 globally
language
89.8
#15 globally
general
46.6
#42 globally
Benchmark Scores
Compare All
Tested on 28 benchmarks · Ranked across 8 categories
Score Distribution (all 312 models)
0255075100
▲ You are here
codingCompare coding →
Aider polyglot

Multi-language code editing from Aider. Tests editing ability across Python, JavaScript, TypeScript, Java, C++, Go, Rust, and more.

47.1·
WeirdML

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

41.0·
Terminal Bench

Complex terminal-based engineering tasks. Models must use command-line tools, navigate filesystems, and debug systems through shell interaction.

17.1·
reasoningCompare reasoning →
HELM — WildBench

Stanford HELM WildBench evaluation. Tests reasoning on challenging real-world tasks.

81.7·
ARC-AGI

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

32.3·
SimpleBench

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

29.4·
mathCompare math →
OTIS Mock AIME 2024-2025

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

73.0·
HELM — Omni-MATH

Stanford HELM evaluation of mathematical reasoning across diverse problem types.

38.4·
FrontierMath-2025-02-28-Private

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

8.5·
Quick compare:Mistral AI logovs Mistral Medium 3z-ai logovs GLM 4 32B Anthropic logovs Claude 2
Excellent (85+) Good (70-85) Average (50-70) Below (<50)
Model Family · Google DeepMind Gemini 2.5 Flash
Jun 2025
37.3
$0.30/M in1.0Mctx28 benchmarks
Jun 2025
$0.15/M in(-0.15)1.0Mctx
Jul 2025
32.5-4.8
$0.10/M in(-0.05)1.0Mctx12 benchmarks
Jul 2025
$0.05/M in(-0.05)1.0Mctx
Sep 2025
$0.10/M in(+0.05)1.0Mctx
See the full Gemini 2.5 Flash family →
Similar Models
Mistral AI logo
Mistral Medium 3
Mistral AI
37.3$0.40/1M
z-ai logo
GLM 4 32B
z-ai
37.5$0.10/1M
Anthropic logo
Claude 2
Anthropic
37.7n/a
Links
Info
Google DeepMindPricing explorerDevelopers · API
Research
Technical Report
Documentation
API DocsPlayground
Community
@Google DeepMind
BenchGecko API
gemini-2-5-flash
Specifications
  • Typemultimodal
  • Context1.0M tokens (~524 books)
  • ReleasedJun 2025
  • LicenseProprietary
  • StatusActive
  • Cost / Message~$0.003
Available On
Google DeepMind logoGoogle DeepMind$0.30
Categories
codingreasoningmathknowledgeagenticlanguagegeneral
Learn More
context-windowtransformertokens
Share & Export
Tweet
Related Models
Mistral AI logoMistral Medium 3
37.3
z-ai logoGLM 4 32B
37.5
Anthropic logoClaude 2
37.7
OpenAI logoGPT-5 Nano
37.7
Gemini 2.5 Flash is a proprietary multimodal AI model by Google DeepMind, released in June 2025. It has an average benchmark score of 37.3. Context window: 1M tokens.

Related Models

Mistral Medium 3 · Mistral AIGLM 4 32B · z-aiClaude 2 · AnthropicGPT-5 Nano · OpenAIHermes 2 Pro - Llama-3 8B · nousresearch

Benchmarks

Chatbot Arena Elo — OverallHELM — IFEvalHELM — WildBenchLech Mazur WritingOTIS Mock AIME 2024-2025

Related Pages

Google DeepMind · ProviderGoogle DeepMind · EconomyAll ModelsCompare ModelsPricingDevelopers · APIContext Window · Glossary

Key facts · as of 2026-10-05

  • Gemini 2.5 Flash by Google DeepMind. BenchGecko score 37.3, rank 210 of 312 scored models (normalized average of public benchmark scores).
  • List price $0.30 input · $2.50 output per 1M tokens (as of 2026-10-05).
  • Sold by 7 providers (as of 2026-10-05): Google AI Studio $0.15 in / $1.25 out · Google $0.30 in / $2.50 out · Google $0.30 in / $2.50 out · Google $0.30 in / $2.50 out · Google AI Studio $0.30 in / $2.50 out · and 2 more. Every provider

How to cite · data as of 2026-10-05

Gemini 2.5 Flash · benchmarks, pricing and providers. BenchGecko, data as of 2026-10-05. https://benchgecko.ai/model/gemini-2-5-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