NEWHow AI models draw the world map · Gecko Tests
Beta
Home/Models/Gemma 4 26B A4B
Google DeepMind logo

Gemma 4 26B A4B

by Google DeepMind · Released Apr 2026

Open SourceMultimodal
47.8
avg score
Rank #159
Compare
Better than 49% of all models
Context
262K tokens (~131 books)
Input $/1M
$0.07
Output $/1M
$0.23
Type
multimodal
License
Open Source
Benchmarks
19 tested
Data as of
About

Gemma 4 26B A4B IT is an instruction-tuned Mixture-of-Experts (MoE) model from Google DeepMind. Despite 25.2B total parameters, only 3.8B activate per token during inference — delivering near-31B quality at...

Tested on 19 benchmarks · BenchGecko score 47.8. Top scores: Chatbot Arena Elo — Overall (1437.5%), Chatbot Arena Elo — Coding (1358.5%), OTIS Mock AIME 2024-2025 (82.2%).

Capabilities
coding
35.2
#153 globally
math
82.2
#21 globally
knowledge
32.7
#235 globally
speed
38.5
#71 globally
general
46.6
#43 globally
Benchmark Scores
Compare All
Tested on 19 benchmarks · Ranked across 6 categories
Score Distribution (all 312 models)
0255075100
▲ You are here
codingCompare coding →
WeirdML

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

35.2·
mathCompare math →
OTIS Mock AIME 2024-2025

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

82.2·
knowledgeCompare knowledge →
GPQA diamond

Graduate-level science questions written by PhD experts. Diamond subset contains questions where experts disagree, testing deep understanding.

64.3·
Chess Puzzles

Tactical chess puzzles testing pattern recognition and multi-move calculation. Measures strategic reasoning ability.

1.1·
Quick compare:Alibaba Qwen logovs Qwen3 32BxAI logovs Grok-3 miniMeta logovs Llama 3.3 70B Instruct
Excellent (85+) Good (70-85) Average (50-70) Below (<50)
Recently Happened
Gemma 4 26B A4B pricing dropped 12%
Oct 3, 2026
Gemma 4 26B A4B pricing increased 13%
Sep 29, 2026
Gemma 4 26B A4B pricing dropped 25%
Sep 26, 2026
Gemma 4 26B A4B pricing increased 114%
Sep 13, 2026
Gemma 4 26B A4B pricing dropped 40%
Sep 11, 2026
Gemma 4 26B A4B pricing dropped 42%
Aug 17, 2026
Similar Models
Alibaba Qwen logo
Qwen3 32B
Alibaba Qwen
47.8$0.08/1M
xAI logo
Grok-3 mini
xAI
47.7n/a
Meta logo
Llama 3.3 70B Instruct
Meta
47.9$0.10/1M
Links
Info
Google DeepMindPricing explorerDevelopers · API
Research
Technical Report
Documentation
API DocsPlayground
Community
@Google DeepMind
Source Code
GitHubHugging Face
BenchGecko API
gemma-4-26b-a4b-it
Specifications
  • Typemultimodal
  • Context262K tokens (~131 books)
  • ReleasedApr 2026
  • LicenseOpen Source
  • StatusActive
  • Cost / Message~$0.000
Available On
Google DeepMind logoGoogle DeepMind$0.07
Categories
codingmathknowledgespeedgeneral
Learn More
context-windowtransformeropen weightstokens
Share & Export
Tweet
Related Models
Alibaba Qwen logoQwen3 32B
47.8
xAI logoGrok-3 mini
47.7
Meta logoLlama 3.3 70B Instruct
47.9
DeepSeek logoR1 Distill Llama 70B
47.3
Gemma 4 26B A4B is an open-source multimodal AI model by Google DeepMind, released in April 2026. It has an average benchmark score of 47.8. Context window: 262K tokens.

Related Models

Qwen3 32B · Alibaba QwenGrok-3 mini · xAILlama 3.3 70B Instruct · MetaR1 Distill Llama 70B · DeepSeekQwen3 4B Thinking 2507 · Alibaba

Benchmarks

Chatbot Arena Elo — OverallChatbot Arena Elo — CodingOTIS Mock AIME 2024-2025Artificial Analysis · GPQA DiamondArtificial Analysis · IFBench

Related Pages

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

Key facts · as of 2026-10-05

  • Gemma 4 26B A4B by Google DeepMind. BenchGecko score 47.8, rank 158 of 312 scored models (normalized average of public benchmark scores).
  • List price $0.0675 input · $0.23 output per 1M tokens (as of 2026-10-05).
  • Sold by 13 providers (as of 2026-10-05): Darkbloom $0.0420 in / $0.22 out · DekaLLM (bf16) $0.0600 in / $0.33 out · DeepInfra (fp8) $0.0700 in / $0.34 out · Makora $0.0800 in / $0.32 out · NextBit (bf16) $0.0900 in / $0.30 out · and 8 more. Every provider
  • Gecko Tests: Who Are You B (Knows who made it) · Tokenizer Tax B (44% more tokens outside English).

How to cite · data as of 2026-10-05

Gemma 4 26B A4B · benchmarks, pricing and providers. BenchGecko, data as of 2026-10-05. https://benchgecko.ai/model/gemma-4-26b-a4b-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