llmboard.ai
Benchmarks
CompareRankings
llmboard.ai
Benchmarks
CompareRankings
HomeModelsGemma 3 27B

Google model product

Gemma 3 27B

Gemma 3 27B is a 27-billion-parameter vision-language model from Google, handling text and image input and generating text output. It features a 128K context window, multilingual support, and open weights. Suitable for complex question answering, summarization, reasoning, and image understanding tasks.

Updated Aug 10, 2026. Default version: Gemma 3 27B

Compare
LLMBoard score45.1Gemma 3 27B
Coverage80%16 benchmark families
Context window131.1KTokens
Official input priceN/AOfficial price unavailable

On this page

  • Specification
  • Capability
  • Benchmarks
  • Arena
  • Pricing
  • Versions
  • About
  • Compare
  • Similar models
  • FAQ

Model specification

Structured fields from the published default version.

Version
Gemma 3 27B
Released
Mar 12, 2025
Knowledge cutoff
Unknown
Parameters
27B
Context window
131.1K
Max output
131.1K
Inputs
image, text
Outputs
text
Open weights
No
License
Gemma

Capability profile

This profile uses the latest version under this unique model that has a calculated LLMBoard score. Arena and price are excluded.

Gemma 3 27B category scores

Benchmark results

Published benchmark records for the scored version Gemma 3 27B.

27 rows
Columns

Show columns

ECLeKTic0.22885.7%CAug 7, 2026
HiddenMath0.621391.7%CAug 7, 2026
VQAv2 (val)0.72350.0%CAug 7, 2026
Bird-SQL (dev)0.53766.7%CAug 7, 2026
Natural2Code0.83871.4%CAug 7, 2026
BIG-Bench Hard0.942185.0%CAug 7, 2026
Global-MMLU-Lite0.861461.5%CAug 7, 2026
InfoVQA0.76937.5%CAug 7, 2026
BIG-Bench Extra Hard0.271140.0%CAug 7, 2026
FACTS Grounding0.771350.0%CAug 7, 2026
MATH0.977191.4%CAug 7, 2026
GSM8k1.084885.1%CAug 7, 2026
MMMU (val)0.691120.0%CAug 7, 2026
IFEval0.9136581.3%CAug 7, 2026
TextVQA0.714157.1%CAug 7, 2026
WMT24++0.5142340.9%CAug 7, 2026
MBPP0.7193750.0%CAug 7, 2026
ChartQA0.8212413.0%CAug 7, 2026
MathVista-Mini0.721239.1%CAug 7, 2026
MRCR v2 (8-needle)0.121210.0%CAug 7, 2026
AI2D0.8233229.0%CAug 7, 2026
DocVQA0.9232612.0%CAug 7, 2026
HumanEval0.9307661.3%CAug 7, 2026
SimpleQA0.1414611.1%CAug 7, 2026
LiveCodeBench0.3627315.3%CAug 7, 2026
MMLU-Pro0.79212928.9%CAug 7, 2026
GPQA0.419723315.5%CAug 7, 2026

Arena results

Preference and agent-evaluation signals from published Arena datasets.

84 rows
Columns

Show columns

visioncreative writing251182.1732N/AJan 9, 2026
vision style controlcreative writing251174.6732N/AJan 9, 2026
visioncaptioning271150.9261N/AAug 6, 2026
vision style controlcaptioning291117.8261N/AAug 6, 2026
visionentity recognition341200.4321N/AAug 6, 2026
vision style controlentity recognition371174.7321N/AAug 6, 2026
vision style controlhumor651152.4770N/AAug 6, 2026
visionhumor661165.9770N/AAug 6, 2026
visioncreative writing vision721184.3626N/AAug 6, 2026
vision style controlcreative writing vision721174.0626N/AAug 6, 2026
visionchinese731154.1728N/AAug 6, 2026
vision style controlchinese741169.0728N/AAug 6, 2026
vision style controlhomework821171.81,050N/AAug 6, 2026
visionhomework841164.81,050N/AAug 6, 2026
visionenglish851186.28,272N/AAug 6, 2026
visiondiagram871157.61,743N/AAug 6, 2026
vision style controlocr881162.46,278N/AAug 6, 2026
visionocr891158.66,278N/AAug 6, 2026
vision style controlenglish901171.18,272N/AAug 6, 2026
visionoverall911165.317,179N/AAug 6, 2026
vision style controloverall911159.717,179N/AAug 6, 2026
vision style controldiagram911166.51,743N/AAug 6, 2026
text factualitynon english1361369.3973N/AAug 6, 2026
text factualityexclude ties1371356.41,662N/AAug 6, 2026
text factualityoverall1381381.72,389N/AAug 6, 2026
textgerman1391362.81,164N/AAug 6, 2026
text factualityhard prompts1391392.71,045N/AAug 6, 2026
text style controljapanese1401309.4951N/AAug 6, 2026
textjapanese1411286.8951N/AAug 6, 2026
text style controlgerman1411371.31,164N/AAug 6, 2026
textkorean1431308.3865N/AAug 6, 2026
text style controlpolish1441368.02,795N/AAug 6, 2026
textindustry medicine and healthcare1451390.12,361N/AAug 6, 2026
text style controlkorean1451324.9865N/AAug 6, 2026
textcreative writing1481344.56,376N/AAug 6, 2026
textfrench1481369.4596N/AAug 6, 2026
textindustry life and physical and social science1511383.17,927N/AAug 6, 2026
textindustry legal and government1531374.62,714N/AAug 6, 2026
textpolish1531344.32,795N/AAug 6, 2026
textindustry business and management and financial operations1541359.66,578N/AAug 6, 2026
text style controlfrench1551387.7596N/AAug 6, 2026
text style controlindustry medicine and healthcare1551403.52,361N/AAug 6, 2026
text style controlcreative writing1561347.56,376N/AAug 6, 2026
textindustry writing and literature and language1591347.010,671N/AAug 6, 2026
text style controlrussian1591368.42,979N/AAug 6, 2026
text style controlindustry business and management and financial operations1621376.86,578N/AAug 6, 2026
textrussian1641349.32,979N/AAug 6, 2026
textspanish1651342.1775N/AAug 6, 2026
text style controlindustry life and physical and social science1651390.67,927N/AAug 6, 2026
text style controlindustry writing and literature and language1661354.310,671N/AAug 6, 2026
text style controlindustry legal and government1681382.02,714N/AAug 6, 2026
textindustry entertainment and sports and media1701320.88,078N/AAug 6, 2026
textoverall1711358.347,453N/AAug 6, 2026
textexclude ties1711322.333,294N/AAug 6, 2026
text style controlnon english1711353.122,650N/AAug 6, 2026
textenglish1721374.324,793N/AAug 6, 2026
text style controlspanish1721350.0775N/AAug 6, 2026
textnon english1731337.022,650N/AAug 6, 2026
text style controlexclude ties1741331.933,294N/AAug 6, 2026
text style controlindustry entertainment and sports and media1741327.78,078N/AAug 6, 2026
text style controloverall1751365.847,453N/AAug 6, 2026
text style controlmulti turn1781358.27,019N/AAug 6, 2026
textmulti turn1791345.27,019N/AAug 6, 2026
text style controllonger query1821363.86,626N/AAug 6, 2026
textinstruction following1831321.612,508N/AAug 6, 2026
textlonger query1841333.06,626N/AAug 6, 2026
text style controlenglish1841373.924,793N/AAug 6, 2026
text style controlinstruction following1851343.812,508N/AAug 6, 2026
text style controlchinese1861373.12,286N/AAug 6, 2026
textchinese1871346.62,286N/AAug 6, 2026
texthard prompts1891339.917,840N/AAug 6, 2026
textindustry software and it services1921355.914,415N/AAug 6, 2026
text style controlhard prompts1921365.617,840N/AAug 6, 2026
texthard prompts english1931349.59,896N/AAug 6, 2026
textindustry mathematical1941326.93,189N/AAug 6, 2026
text style controlindustry mathematical1951340.13,189N/AAug 6, 2026
textexpert1981302.42,221N/AAug 6, 2026
text style controlhard prompts english1991370.39,896N/AAug 6, 2026
text style controlindustry software and it services2001377.914,415N/AAug 6, 2026
textmath2011311.43,576N/AAug 6, 2026
text style controlexpert2011335.12,221N/AAug 6, 2026
text style controlmath2011322.33,576N/AAug 6, 2026
textcoding2081322.78,067N/AAug 6, 2026
text style controlcoding2301358.08,067N/AAug 6, 2026

Pricing

Official vendor API PAYG pricing is summarized first. The table then lists individual provider offerings without treating their minimum as the official price.

Official API
N/A
Official provider
N/A
Lowest third-party
From $0.08 input, $0.45 output per 1M via OpenRouter
Tracked offerings
6
6 rows
Columns

Show columns

OpenRoutergoogle/gemma-3-27b-itglobal$0.08$0.45262.1KAug 7, 2026
Cortecsgemma-3-27b-itglobal$0.099$0.299131KAug 7, 2026
LLM Gatewaygemma-3-27bglobal$0.10$0.30110KAug 7, 2026
NovitaAIgoogle/gemma-3-27b-itglobal$0.119$0.2098.3KAug 7, 2026
GreenPTgemma-3-27b-itglobal$0.342$0.68440KAug 7, 2026
STACKITgoogle/gemma-3-27b-itglobal$0.53$0.7637KAug 7, 2026

Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.

Gemma 3 27B versions

All published versions linked to this unique model. The score columns identify the version used by the current overall ranking.

1 rows
Columns

Show columns

Gemma 3 27BMar 12, 202545.127B131.1K131.1KNoGemma

What is Gemma 3 27B?

A concise description based on the published model registry.

Gemma 3 27B is a 27-billion-parameter vision-language model from Google, handling text and image input and generating text output. It features a 128K context window, multilingual support, and open weights. Suitable for complex question answering, summarization, reasoning, and image understanding tasks.

Use the benchmark, Arena and pricing sections above as separate evidence. A missing field means the current data snapshot does not support that claim.

Data snapshot: 2026-08-07. Editorial model content is not available in the backend.

Gemma 3 27B vs nearby models

Open a comparison with the three ranked models immediately above and below this model.

Gemma 3 27BvsGemma 2 27BGemma 3 27BvsSarvam 30BGemma 3 27BvsQwen2.5 Coder 32BGemma 3 27BvsNemotron 3 NanoGemma 3 27BvsGranite 3.3 8BGemma 3 27BvsMiniMax M2

Models similar to Gemma 3 27B

Recommendations prioritize the same model type and family, then the closest published LLMBoard score.

#93+1.9
GO

Gemma 2 27B

Google

47.1 LLMBoard

DetailsCompare
#91+3.0
GO

Gemma 4 12B

Google

48.1 LLMBoard

DetailsCompare
#87+3.4
GO

Gemini 1.5 Pro

Google

48.6 LLMBoard

DetailsCompare
#116-6.0
GO

Gemma 3 12B

Google

39.2 LLMBoard

DetailsCompare
#118-6.9
GO

DiffusionGemma 26B A4B

Google

38.3 LLMBoard

DetailsCompare
#76+7.5
GO

Gemini 2.0 Flash Lite

Google

52.6 LLMBoard

DetailsCompare

FAQ

Common questions about Gemma 3 27B.

When was Gemma 3 27B released?

Gemma 3 27B's default version was released on Mar 12, 2025.

How much does Gemma 3 27B cost?

No official standard PAYG price is currently available for Gemma 3 27B. The lowest tracked third-party offer starts at $0.08 input and $0.45 output via OpenRouter.

Who created Gemma 3 27B?

Gemma 3 27B is published under Google in the model registry.

What is the context window for Gemma 3 27B?

The default version has a 131.1K token context window.

Is Gemma 3 27B open weight?

No. The default version is not marked as having publicly available weights.

How many API providers offer Gemma 3 27B?

6 published provider offerings are linked to the default version.

What models should I compare Gemma 3 27B with?

Nearby ranked alternatives include Gemma 2 27B, Sarvam 30B, Qwen2.5 Coder 32B.

Rankings

OverallCodingText ArenaPricing

Modalities

Image GenerationVideo GenerationSpeech-to-TextEmbeddings

Benchmarks

All BenchmarksReasoningMathCoding

Vendors

All VendorsOpenAIAnthropicGoogle
llmboard.aiCopyright 2026 llmboard.ai