llmboard.ai
Benchmarks
CompareRankings
llmboard.ai
Benchmarks
CompareRankings
HomeModelsGemma 4 26B A4B

Google model product

Gemma 4 26B A4B

Gemma 4 26B-A4B is Google DeepMind's Mixture-of-Experts multimodal model with 26 billion total parameters and 3.8 billion activated parameters per token, with a 256K context window. Ranks #6 among open models on Arena AI, outcompeting models 20x its size. Achieves an estimated LMArena text score of 1441. Supports image and text inputs.

Updated Aug 10, 2026. Default version: Gemma 4 26B-A4B

Compare
LLMBoard score67.2Gemma 4 26B-A4B
Coverage60%7 benchmark families
Context window262.1KTokens
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 4 26B-A4B
Released
Apr 2, 2026
Knowledge cutoff
Jan 1, 2025
Parameters
25.2B
Context window
262.1K
Max output
32.8K
Inputs
image, text
Outputs
text
Open weights
Yes
License
Apache 2.0

Capability profile

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

Gemma 4 26B-A4B category scores

Benchmark results

Published benchmark records for the scored version Gemma 4 26B-A4B.

12 rows
Columns

Show columns

BIG-Bench Extra Hard0.621190.0%CAug 7, 2026
MedXpertQA0.661254.5%CAug 7, 2026
t2-bench0.962377.3%CAug 7, 2026
MRCR v2 (8-needle)0.482165.0%CAug 7, 2026
AIME 20260.9121731.3%CAug 7, 2026
MathVision0.8123264.5%CAug 7, 2026
LiveCodeBench v60.8235357.7%CAug 7, 2026
MMMLU0.9244952.1%CAug 7, 2026
MMLU-Pro0.83412974.2%CAug 7, 2026
MMMU-Pro0.7356546.9%CAug 7, 2026
Humanity's Last Exam0.2649230.8%CAug 7, 2026
GPQA0.87323369.0%CAug 7, 2026

Arena results

Preference and agent-evaluation signals from published Arena datasets.

82 rows
Columns

Show columns

visioncaptioning41300.743N/AAug 6, 2026
vision style controlcaptioning51247.843N/AAug 6, 2026
visionhumor221278.6877N/AAug 6, 2026
visionhomework291292.93,614N/AAug 6, 2026
vision style controlhumor291251.9877N/AAug 6, 2026
vision style controlhomework311284.93,614N/AAug 6, 2026
textmath331467.2370N/AAug 6, 2026
visionchinese331315.71,350N/AAug 6, 2026
visioncreative writing vision341260.91,328N/AAug 6, 2026
vision style controlchinese361294.81,350N/AAug 6, 2026
textindustry mathematical371470.0292N/AAug 6, 2026
text style controlindustry mathematical371475.3292N/AAug 6, 2026
visionocr391266.415,709N/AAug 6, 2026
textfrench401466.2216N/AAug 6, 2026
vision style controlentity recognition401152.8205N/AAug 6, 2026
text style controlmath411467.6370N/AAug 6, 2026
visionoverall411257.822,246N/AAug 6, 2026
visionentity recognition411182.1205N/AAug 6, 2026
vision style controlcreative writing vision441235.31,328N/AAug 6, 2026
text style controlfrench451477.5216N/AAug 6, 2026
vision style controlocr451255.815,709N/AAug 6, 2026
textindustry legal and government461456.2382N/AAug 6, 2026
visiondiagram461263.35,777N/AAug 6, 2026
visionenglish471254.49,046N/AAug 6, 2026
vision style controloverall481239.622,246N/AAug 6, 2026
textindustry business and management and financial operations501434.71,042N/AAug 6, 2026
textrussian501443.3678N/AAug 6, 2026
vision style controldiagram511260.65,777N/AAug 6, 2026
vision style controlenglish511236.89,046N/AAug 6, 2026
text factualityinstruction following551442.71,588N/AAug 6, 2026
text style controlindustry legal and government561462.4382N/AAug 6, 2026
text factualityexclude ties591447.33,943N/AAug 6, 2026
text style controlindustry business and management and financial operations601449.51,042N/AAug 6, 2026
textchinese611477.8389N/AAug 6, 2026
textinstruction following621422.51,590N/AAug 6, 2026
text style controlinstruction following621439.81,590N/AAug 6, 2026
textexpert631448.3405N/AAug 6, 2026
text factualityoverall631445.15,753N/AAug 6, 2026
text factualityenglish631454.52,492N/AAug 6, 2026
text style controlexpert631475.3405N/AAug 6, 2026
textindustry entertainment and sports and media651406.21,037N/AAug 6, 2026
text factualitynon english651433.23,261N/AAug 6, 2026
text style controlrussian651446.8678N/AAug 6, 2026
textnon english671423.73,262N/AAug 6, 2026
text factualityhard prompts english671473.31,469N/AAug 6, 2026
webdevwebdev-react681358.51,220N/AAug 6, 2026
textenglish691442.52,493N/AAug 6, 2026
text style controlhard prompts english691469.91,470N/AAug 6, 2026
textexclude ties701429.83,945N/AAug 6, 2026
text style controllonger query711448.51,553N/AAug 6, 2026
webdevwebdev-html711375.0190N/AAug 6, 2026
textoverall721434.75,755N/AAug 6, 2026
text style controlhard prompts721462.33,245N/AAug 6, 2026
textcreative writing731404.5939N/AAug 6, 2026
textlonger query731429.41,553N/AAug 6, 2026
text factualityhard prompts731463.03,243N/AAug 6, 2026
text style controlindustry entertainment and sports and media731410.31,037N/AAug 6, 2026
webdevoverall731366.51,415N/AAug 6, 2026
webdevwebdev731366.51,415N/AAug 6, 2026
texthard prompts english741446.11,470N/AAug 6, 2026
textindustry software and it services741453.12,077N/AAug 6, 2026
textmulti turn741438.91,082N/AAug 6, 2026
text factualityindustry software and it services741476.82,077N/AAug 6, 2026
texthard prompts761439.93,245N/AAug 6, 2026
text style controlenglish761449.22,493N/AAug 6, 2026
text style controlindustry software and it services761474.82,077N/AAug 6, 2026
textindustry writing and literature and language771409.11,303N/AAug 6, 2026
text factualitylonger query771448.21,553N/AAug 6, 2026
text factualitycoding781484.41,357N/AAug 6, 2026
text style controloverall781438.25,755N/AAug 6, 2026
text style controlexclude ties781435.33,945N/AAug 6, 2026
textindustry life and physical and social science801442.3865N/AAug 6, 2026
text style controlmulti turn801446.11,082N/AAug 6, 2026
text style controlnon english821424.63,262N/AAug 6, 2026
text style controlchinese851475.0389N/AAug 6, 2026
text style controlcoding851480.61,357N/AAug 6, 2026
text style controlcreative writing861403.0939N/AAug 6, 2026
text style controlindustry writing and literature and language861409.71,303N/AAug 6, 2026
textcoding881444.61,357N/AAug 6, 2026
text style controlindustry life and physical and social science891450.7865N/AAug 6, 2026
textindustry medicine and healthcare1081423.0364N/AAug 6, 2026
text style controlindustry medicine and healthcare1261431.4364N/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 $0.05 input, $0.29 output per 1M via EmpirioLabs AI
Tracked offerings
5
5 rows
Columns

Show columns

EmpirioLabs AIgemma-4-26b-a4bglobal$0.05$0.29262.1KAug 7, 2026
LLM Gatewaygemma-4-26b-a4b-itglobal$0.07$0.34262.1KAug 7, 2026
Cortecsgemma-4-26b-a4b-itglobal$0.111$0.557262KAug 7, 2026
Charm Hypergemma-4-26b-a4b-itglobal$0.124$0.42256KAug 7, 2026
Scalewaygemma-4-26b-a4b-itglobal$0.25$0.50256KAug 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 4 26B A4B 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 4 26B-A4BApr 2, 202667.225.2B262.1K32.8KYesApache 2.0

What is Gemma 4 26B A4B?

A concise description based on the published model registry.

Gemma 4 26B-A4B is Google DeepMind's Mixture-of-Experts multimodal model with 26 billion total parameters and 3.8 billion activated parameters per token, with a 256K context window. Ranks #6 among open models on Arena AI, outcompeting models 20x its size. Achieves an estimated LMArena text score of 1441. Supports image and text inputs.

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 4 26B A4B vs nearby models

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

Gemma 4 26B A4BvsClaude Opus 4.6Gemma 4 26B A4BvsGPT-5.2Gemma 4 26B A4BvsMiMo V2.5 ProGemma 4 26B A4BvsGrok 4 FastGemma 4 26B A4BvsGLM 4.7Gemma 4 26B A4BvsQwen3 235B A22B Thinking

Models similar to Gemma 4 26B A4B

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

#42-3.7
GO

Gemini 2.5 Flash

Google

63.5 LLMBoard

DetailsCompare
#24+5.4
GO

Gemini 3 Flash

Google

72.6 LLMBoard

DetailsCompare
#21+7.1
GO

Gemma 4 31B

Google

74.3 LLMBoard

DetailsCompare
#14+10.8
GO

Gemini 2.5 Pro

Google

78.0 LLMBoard

DetailsCompare
#73-14.1
GO

Gemini 2.0 Flash

Google

53.1 LLMBoard

DetailsCompare
#76-14.6
GO

Gemini 2.0 Flash Lite

Google

52.6 LLMBoard

DetailsCompare

FAQ

Common questions about Gemma 4 26B A4B.

When was Gemma 4 26B A4B released?

Gemma 4 26B A4B's default version was released on Apr 2, 2026.

How much does Gemma 4 26B A4B cost?

No official standard PAYG price is currently available for Gemma 4 26B A4B. The lowest tracked third-party offer starts at $0.05 input and $0.29 output via EmpirioLabs AI.

Who created Gemma 4 26B A4B?

Gemma 4 26B A4B is published under Google in the model registry.

What is the context window for Gemma 4 26B A4B?

The default version has a 262.1K token context window.

Is Gemma 4 26B A4B open weight?

Yes. The default version is marked as open weight under Apache 2.0.

How many API providers offer Gemma 4 26B A4B?

5 published provider offerings are linked to the default version.

What models should I compare Gemma 4 26B A4B with?

Nearby ranked alternatives include Claude Opus 4.6, GPT-5.2, MiMo V2.5 Pro.

Rankings

OverallCodingText ArenaPricing

Modalities

Image GenerationVideo GenerationSpeech-to-TextEmbeddings

Benchmarks

All BenchmarksReasoningMathCoding

Vendors

All VendorsOpenAIAnthropicGoogle
llmboard.aiCopyright 2026 llmboard.ai