Skip to content

Google · Chat

Gemma 4 26B A4B

Google 🇺🇸ChatGemini familyOpen weights. One of its sources says the weights are publicly available.
From Empiriolabs's listing:
Gemma 4 26B A4B is a Google open multimodal model with 256K context, text, image, and video input, tools, and structured output.
llminfo IQ
124.9
Max context
262K
Listings
114

Evidence and sources

🤗 google/gemma-4-26B-A4B-it Model architecture, as its sources state it llminfo IQ 124.9, ranked among chat models. From 216 benchmarks. Plenty of data behind it.How llminfo IQ is computedCoding IQ 101.7, 95% interval 94–110. From 10 benchmark families across 7 sources. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ.How Coding IQ is computedGood value (56.4). Cheaper per task than its benchmark results predict. From 7 benchmarks with measured task costs. A fair amount of data behind it.How Value for money worksYap score: tokens used per coding task vs. models with the same resultsTypical (50 ±9). Generates about as many tokens as models with similar coding results. Measured at default effort. Little data behind it, so treat it as a rough estimate.Median 22,850 output tokens per task at default effort.How the Yap score worksReleased 31 Mar 2026Maker HQ: US (according to Google's own website)
Sources: LMArena data, EmpirioLabs data, Models.dev data, Morpheus Inference Marketplace data, llminfo researchPrimary data from LMArena from EmpirioLabs: primary mode, description from Models.dev: weights availability from Morpheus Inference Marketplace: architecture from llminfo research: release date
Tags from its sources:open-sourcemultimodalvisionmoelong-contextfunction-callingreasoningmultilingual

Advantages and trade-offs

Compared with 10 chat models, picked mainly by popularity and IQ

Advantages

Open weights available

The weights are public. 3 of the 8 peers that say either way have open weights.

We only make a claim when there's enough data on models of the same type. Updated 22 Nov, 06:07 UTC.Prices come from the model's maker where they sell it, otherwise their official partner, otherwise the market median. This comparison uses 8 maker's prices; 2 market median prices.

Input & output

as stated by the model's providers
Text
Images
Video
Files
Text

Capabilities

Supported operations and provider-stated features

VisionReasoningFunction callingStructured outputPrompt cachingWeb searchBatch API

Listings

 Googlemodels/gemma-4-26b-a4b-itSources: GenAI Prices data, Google Generative AI API data docs, LiteLLM data, Models.dev data262K$0.06$0.33——
 Aihubmixgemma-4-26b-a4b-it-free · freeSource: AIHubMix data docs262K$0.00$0.00——
 Flowdotgoogle/gemma-4-26b-a4b-it:free · freeSource: FlowDot data262K$0.00$0.00——
 OpenRouterInference by Googlegoogle/gemma-4-26b-a4b-it:free · freeSource: OpenRouter data docs262K$0.00$0.00411,007.5ms
 NanoGPTgoogle/gemma-4-26b-a4b-itSources: Models.dev data, NanoGPT262K$0.0354cache read $0.0177$0.0943——
 Kilogoogle/gemma-4-26b-a4b-itSources: Kilo AI Gateway data, Models.dev data262K$0.05cache read $0.0125$0.19——
 Empiriolabsgemma-4-26b-a4b · FP8 weights + FP8 KVSources: EmpirioLabs data, LiteLLM data262K$0.05cache read $0.025$0.29——
 Hugging FaceInference by Dekallmgoogle/gemma-4-26B-A4B-itSource: Hugging Face Inference Providers data262K$0.06$0.33——
 Orcaroutergoogle/gemma-4-26b-a4b-itSource: Models.dev data262K$0.06cache read $0.0075$0.33——
 Llmgateway Providersflexai/gemma-4-26b-a4b-itSource: Models.dev data262K$0.065cache read $0.0097$0.2813——

13 listings without a price are hidden.

Benchmarks

Named effort levels run from lowest to highest. Select a result to see its source details.

Human preference

Arena ELO41 related benchmarks · 41 results

Coding

Arena ELOSource: LMArena data · Nov 2026

Multilingual

Bangla LLM BenchSource: Bangla LLM Bench data docs

General

Instruction following

Knowledge

Math

OCR

Reasoning

Science

Data credits: Epoch AI Benchmarking Hub, CC BY 4.0 (licence) · LMArena, CC BY 4.0 (licence) · Bangla LLM Bench, Md Deluair Hossen. Source tasks: Hendrycks et al. 2021 (MMLU), OpenAI MMMLU, Bandarkar et al. 2024 (Belebele), Shi et al. 2023 (MGSM), Nahin et al. 2025 (TituLLMs). Public numeric community results; results licence not separately stated. · Baromètre de l’écriture IA, Johannes Eva; public numeric ranking data, © 2026 Johannes Eva, no results licence stated · Clinical Benchmarks (clinicalbenchmarks.ai), CC BY 4.0; fractions converted to percentages; source documents retain their own terms (licence) · SnakeBench (Greg Kamradt / ARC Prize Foundation), MIT (licence)

Around the web

  • Gemma explained: An overview of Gemma model family architectures - Google Developers Blog
    ## Gemma Gemma is an open weight LLM. It comes in both instruction-tuned and raw, …## Gemma Gemma is an open weight LLM. It comes in both instruction-tuned and raw, pretrained variants at various parameter sizes. It is based on the LLM architecture introduced by Google Research in …

    developers.googleblog.com

    Docs
  • Gemma 4  |  Google AI Edge  |  Google for Developers
    Gemma 4 models are designed to deliver frontier-level performance at each size, targeting deployment scenarios …Gemma 4 models are designed to deliver frontier-level performance at each size, targeting deployment scenarios from mobile and edge devices (E2B, E4B) to consumer GPUs and workstations (26B A4B, 31B)…

    developers.google.com

    Docs
  • Google Gemma · GitHub
    Gemma is a family of lightweight, state-of-the-art open models built from the same research and …Gemma is a family of lightweight, state-of-the-art open models built from the same research and technology used to create the Gemini models. Developed by Google DeepMind and other teams across Google…

    github.com

    Repository
  • GitHub - google-deepmind/gemma: Gemma open-weight LLM library, from Google DeepMind · GitHub
    ## Repository files navigation # Gemma Unittests PyPI version Documentation Status Gemma is a family …## Repository files navigation # Gemma Unittests PyPI version Documentation Status Gemma is a family of open-weights Large Language Model (LLM) by Google DeepMind, based on Gemini research and techno…

    github.com

    Repository
  • Google | Gemma | Kaggle
    Gemma is a family of lightweight, state-of-the-art open models from Google, built from the same …Gemma is a family of lightweight, state-of-the-art open models from Google, built from the same research and technology used to create the Gemini models. They are text-to-text, decoder-only large lan…

    kaggle.com

  • Gemma: Google introduces new state-of-the-art open models
    Gemma is a family of lightweight, state-of-the-art open models built from the same research and …Gemma is a family of lightweight, state-of-the-art open models built from the same research and technology used to create the Gemini models. Developed by Google DeepMind and other teams across Google…

    blog.google

  • Gemma (language model) - Wikipedia
    In February 2024, Google debuted Gemma, a collection of open weight LLMs that serve as …In February 2024, Google debuted Gemma, a collection of open weight LLMs that serve as a lightweight version of Gemini. The initial release came in two sizes: neural networks with two and seven billi…

    en.wikipedia.org

  • Gemma models overview  |  Google AI for Developers
    Gemma is a family of lightweight, state-of-the-art open models built from the same research and …Gemma is a family of lightweight, state-of-the-art open models built from the same research and technology used to create the Gemini models. Developed by Google DeepMind and other teams across Google…

    ai.google.dev

  • What is Gemma? Google’s open sourced AI model explained | TechTarget
    The name Gemma comes from the Latin word for precious stone. Google released Gemma on …The name Gemma comes from the Latin word for precious stone. Google released Gemma on Feb. 21, 2024, with two models: Gemma 2B and Gemma 7B. Gemma has had multiple versions and iterations. Google fir…

    techtarget.com

  • What Is Google Gemma? | IBM
    These text-to-text artificial intelligence (AI) models derive their name from the same Latin word, which …These text-to-text artificial intelligence (AI) models derive their name from the same Latin word, which means “precious stone.” Gemma is a group of open models, with Google providing free access to …

    ibm.com

Also known as: @cf/google/gemma-4-26b-a4b-it, aihubmix/gemma-4-26b-a4b-it, bedrock_mantle/eu-central-1/google.gemma-4-26b-a4b, bedrock_mantle/google.gemma-4-26b-a4b, bedrock_mantle/us-gov-west-1/google.gemma-4-26b-a4b, cloudflare/@cf/google/gemma-4-26b-a4b-it, deepinfra-gemma-4-26b-a4b-it, deepinfra/gemma-4-26b-a4b-it, deepinfra/google/gemma-4-26b-a4b-it, empiriolabs/gemma-4-26b-a4b, flexai/gemma-4-26b-a4b-it, gemini/gemma-4-26b-a4b-it, gemma-4-26b-a4b-instruct, gemma-4-26b-a4b-it, gemma-4-26b-a4b-it-free, gemma-4-26b-a4b:web, gmi/google/gemma-4-26b-a4b-it, google gemma 4 26b a4b instruct, google-gemma-4-26b-a4b-it, google.gemma-4-26b-a4b, google/gemma-4-26b-a4b, google/gemma-4-26b-a4b-it, google/gemma-4-26b-a4b-it-maas, google/gemma-4-26b-a4b-it:free, google/gemma-4-26b-a4b-it:thinking, novita/gemma-4-26b-a4b-it, novita/google/gemma-4-26b-a4b-it, parasail-gemma-4-26b-a4b-it, parasail/google/gemma-4-26b-a4b-it, scaleway/google/gemma-4-26b-a4b-it, vertex_ai/gemma-4-26b-a4b-it, vertex_ai/google/gemma-4-26b-a4b-it-maas, wandb/google/gemma-4-26b-a4b-it

Like this one. Tap or hover any dotted value, label or icon to see where it comes from and how it is measured. open details when tapped or hovered. How this data works