NEWHow AI models draw the world map · Gecko Tests
Beta
Home/Models/Gemma 3 27B
Google DeepMind logo

Gemma 3 27B

by Google DeepMind · Released Mar 2025

Open SourceMultimodal
22.7
avg score
Rank #272
Compare
Better than 13% of all models
Context
131K tokens (~66 books)
Input $/1M
$0.08
Output $/1M
$0.45
Type
multimodal
License
Open Source
Benchmarks
17 tested
Data as of
About

Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities,...

Tested on 17 benchmarks · BenchGecko score 22.7. Top scores: Chatbot Arena Elo — Overall (1365.4%), OpenCompass — IFEval (81.0%), Lech Mazur Writing (79.9%).

Capabilities
coding
17.9
#180 globally
math
39.6
#156 globally
knowledge
39.4
#207 globally
general
17.7
#172 globally
language
81.0
#57 globally
Benchmark Scores
Compare All
Tested on 17 benchmarks · Ranked across 6 categories
Score Distribution (all 312 models)
0255075100
▲ You are here
codingCompare coding →
OpenCompass — LiveCodeBenchV6

OpenCompass Live Code Bench v6. Fresh competitive programming problems to evaluate code generation without memorization.

30.8·
Aider polyglot

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

4.9·
mathCompare math →
MATH level 5

Competition-level math from AMC, AIME, and olympiad problems. Level 5 is the hardest tier, requiring creative problem-solving.

74.0·
OTIS Mock AIME 2024-2025

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

22.4·
OpenCompass — AIME2025

OpenCompass evaluation on AIME 2025 problems. Tests mathematical reasoning on fresh competition problems.

22.4·
knowledgeCompare knowledge →
Lech Mazur Writing

Writing quality evaluation by Lech Mazur. Tests prose quality, coherence, and stylistic ability.

79.9·
OpenCompass — MMLU-Pro

OpenCompass MMLU-Pro evaluation. Harder knowledge test with more answer choices.

67.8·
GeoBench

Geography benchmark testing knowledge of world geography, landmarks, borders, and geopolitical facts.

52.0·
Quick compare:Mistral AI logovs Mistral Small 3.1 24BOpenAI logovs GPT-4.1 NanoAnthropic logovs Claude 2.1
Excellent (85+) Good (70-85) Average (50-70) Below (<50)
Similar Models
Mistral AI logo
Mistral Small 3.1 24B
Mistral AI
22.7$0.35/1M
OpenAI logo
GPT-4.1 Nano
OpenAI
22.4$0.10/1M
Anthropic logo
Claude 2.1
Anthropic
22.2n/a
Links
Info
Google DeepMindPricing explorerDevelopers · API
Research
Technical Report
Documentation
API DocsPlayground
Community
@Google DeepMind
Source Code
GitHubHugging Face
BenchGecko API
gemma-3-27b-it
Specifications
  • Typemultimodal
  • Context131K tokens (~66 books)
  • ReleasedMar 2025
  • LicenseOpen Source
  • StatusActive
  • Cost / Message~$0.001
Available On
Google DeepMind logoGoogle DeepMind$0.08
Categories
codingmathknowledgegenerallanguage
Learn More
context-windowtransformeropen weightstokens
Share & Export
Tweet
Related Models
Mistral AI logoMistral Small 3.1 24B
22.7
OpenAI logoGPT-4.1 Nano
22.4
Anthropic logoClaude 2.1
22.2
Meta logoLlama 4 Maverick
22.1
Gemma 3 27B is an open-source multimodal AI model by Google DeepMind, released in March 2025. It has an average benchmark score of 22.7. Context window: 131K tokens.

Related Models

Mistral Small 3.1 24B · Mistral AIGPT-4.1 Nano · OpenAIClaude 2.1 · AnthropicLlama 4 Maverick · MetaMistral Small 3.2 24B · Mistral AI

Benchmarks

Chatbot Arena Elo — OverallOpenCompass — IFEvalLech Mazur WritingMATH level 5OpenCompass — MMLU-Pro

Related Pages

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

Key facts · as of 2026-10-05

  • Gemma 3 27B by Google DeepMind. BenchGecko score 22.7, rank 272 of 312 scored models (normalized average of public benchmark scores).
  • List price $0.0800 input · $0.45 output per 1M tokens (as of 2026-10-05).
  • Sold by 4 providers (as of 2026-10-05): DeepInfra (fp8) $0.0800 in / $0.16 out · Parasail (fp8) $0.0800 in / $0.45 out · Nebius (fp8) $0.10 in / $0.30 out · Novita (bf16) $0.12 in / $0.20 out. Every provider

How to cite · data as of 2026-10-05

Gemma 3 27B · benchmarks, pricing and providers. BenchGecko, data as of 2026-10-05. https://benchgecko.ai/model/gemma-3-27b-it

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