Gemma 4 26B A4B: Specifications & Sources

Specifications and access

Specification	Value and source
Weights	Open weightsSource
Access	API, LocalSource
Parameters	25.2BSource
Active parameters	3.8BSource
Context window	262,144 tokensSource
Knowledge cutoff	January 2025Source
Release	2026-03-31Source
Parameters	25.2 BSource
Input	Text, ImageSource
Output	TextSource
Architecture	Mixture-of-experts transformer, 25.2B total and 3.8B active; hybrid sliding/global attentionSource
License	Apache 2.0Source
Knowledge cutoff	2025-01Source
Active parameters	3.8 BSource
Languages	140+Source
Announcement	2026-04-02Source

Measurements without matching peer values

These measurements have no matching peer values under the same test conditions. Their original values and sources remain available here.

MMLU Pro (Show measurement, test conditions, and source)

Source value: 82.6
Score: 82.6
Metric: percent
Unit: %
Category: vendor-evaluation
Direction: Higher is better
Source type: vendor-reported
Evaluator: Google DeepMind
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Gemma 4 model-card benchmark table; 26B A4B column

Source Google DeepMind

AIME 2026 (no tools) (Show measurement, test conditions, and source)

Source value: 88.3
Score: 88.3
Metric: percent
Unit: %
Category: vendor-evaluation
Direction: Higher is better
Source type: vendor-reported
Evaluator: Google DeepMind
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Gemma 4 model-card benchmark table; 26B A4B column

Source Google DeepMind

LiveCodeBench v6 (Show measurement, test conditions, and source)

Source value: 77.1
Score: 77.1
Metric: percent
Unit: %
Category: vendor-evaluation
Direction: Higher is better
Source type: vendor-reported
Evaluator: Google DeepMind
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Gemma 4 model-card benchmark table; 26B A4B column

Source Google DeepMind

GPQA (Show measurement, test conditions, and source)

Source value: 82.3
Score: 82.3
Metric: percent
Unit: %
Category: vendor-evaluation
Direction: Higher is better
Source type: vendor-reported
Evaluator: Google DeepMind
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Gemma 4 model-card benchmark table; 26B A4B column

Source Google DeepMind

Tau2 (Show measurement, test conditions, and source)

Source value: 68.2
Score: 68.2
Metric: percent
Unit: %
Category: vendor-evaluation
Direction: Higher is better
Source type: vendor-reported
Evaluator: Google DeepMind
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Gemma 4 model-card benchmark table; 26B A4B column

Source Google DeepMind

Humanity's Last Exam (no tools) (Show measurement, test conditions, and source)

Source value: 8.7
Score: 8.7
Metric: percent
Unit: %
Category: vendor-evaluation
Direction: Higher is better
Source type: vendor-reported
Evaluator: Google DeepMind
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Gemma 4 model-card benchmark table; 26B A4B column

Source Google DeepMind

Humanity's Last Exam (search) (Show measurement, test conditions, and source)

Source value: 17.2
Score: 17.2
Metric: percent
Unit: %
Category: vendor-evaluation
Direction: Higher is better
Source type: vendor-reported
Evaluator: Google DeepMind
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Gemma 4 model-card benchmark table; 26B A4B column

Source Google DeepMind

BigBench Extra Hard (Show measurement, test conditions, and source)

Source value: 64.8
Score: 64.8
Metric: percent
Unit: %
Category: vendor-evaluation
Direction: Higher is better
Source type: vendor-reported
Evaluator: Google DeepMind
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Gemma 4 model-card benchmark table; 26B A4B column

Source Google DeepMind

MMMLU (Show measurement, test conditions, and source)

Source value: 86.3
Score: 86.3
Metric: percent
Unit: %
Category: vendor-evaluation
Direction: Higher is better
Source type: vendor-reported
Evaluator: Google DeepMind
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Gemma 4 model-card benchmark table; 26B A4B column

Source Google DeepMind

MMMU Pro (Show measurement, test conditions, and source)

Source value: 73.8
Score: 73.8
Metric: percent
Unit: %
Category: vendor-evaluation
Direction: Higher is better
Source type: vendor-reported
Evaluator: Google DeepMind
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Gemma 4 model-card benchmark table; 26B A4B column

Source Google DeepMind

MATH-Vision (Show measurement, test conditions, and source)

Source value: 82.4
Score: 82.4
Metric: percent
Unit: %
Category: vendor-evaluation
Direction: Higher is better
Source type: vendor-reported
Evaluator: Google DeepMind
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Gemma 4 model-card benchmark table; 26B A4B column

Source Google DeepMind

MedXpertQA (Show measurement, test conditions, and source)

Source value: 58.1
Score: 58.1
Metric: percent
Unit: %
Category: vendor-evaluation
Direction: Higher is better
Source type: vendor-reported
Evaluator: Google DeepMind
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Gemma 4 model-card benchmark table; 26B A4B column

Source Google DeepMind

Page 1 of 2

Our tests

Each test shows the latest attempt, with no subsequent fixes to the generated code. Earlier runs remain available in the test details. These results do not contribute to an overall score.

Pelican on a bicycle

A community classic for free-form SVG drawing.

Gemma 4 26B A4B

Reasoning (requested): high

Current attempt

Run date: 1 Oct 2026
API provider: openrouter
Requested API model: google/gemma-4-26b-a4b-it
API model reported in response: google/gemma-4-26b-a4b-it
Reasoning setting: high
Pinned endpoint: Darkbloom | google/gemma-4-26b-a4b-it-20260403
Temperature: Provider default
Token limit including reasoning: 8,192 tokens
Test protocol: community-visual-tests-v1
Technical checks: Passed, not a quality rating
Response time: 86.1 s

Visually reviewed. This individual test does not establish overall model quality.

Prompt and test conditions

Generate an SVG of a pelican riding a bicycle

Token limit including reasoning: 8,192 tokens

Task origin (Simon Willison)

Popularity

49.4 / 100

Published prices

Hosting-market price: the median of 11 OpenRouter endpoints from 11 providers (representative: Makora), observed on October 7, 2026. Cache prices differ by provider and are not shown.

API input: $0.08 per 1M tokensSource
API output: $0.32 per 1M tokensSource

Sources and data date

Type	Evidence and data date
Price	Base API prices without caching or batch discounts. Higher context tiers and other rates are listed under Costs. Source: Pricing · API pricing
Input	Documented modalities. Retrieved September 8, 2026. Source: Model documentation · Knowledge cutoff data (retrieved September 8, 2026) · Gemma 4 model card · Editorial description reviewed October 2, 2026
Output	Documented modalities. Retrieved September 8, 2026. Source: Model documentation · Knowledge cutoff data (retrieved September 8, 2026) · Gemma 4 model card · Editorial description reviewed October 2, 2026
Research status	Research date September 8, 2026. 4 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication.
Not found in the inspected sources	active-parameters-billions Original wording Active parameter count is not separately published for this dense model.
Not found in the inspected sources	api-model-id Original wording Gemma is distributed as open weights; the card does not define a hosted Gemini API model ID.
Not found in the inspected sources	training-tokens Original wording The Gemma 4 model card describes the dataset but does not publish a token-count total.
Additional source	Model metadata · Gemma 4 26B A4B instruction-tuned weights · Editorial description reviewed October 2, 2026
Additional source	Artificial Analysis (retrieved September 8, 2026)
Additional source	Model metadata (retrieved September 8, 2026)
Additional source	Model metadata (retrieved September 8, 2026)
Additional source	OpenRouter: Popularity index (derived), CC BY 4.0
Additional source	Vercel AI Gateway: Popularity index (derived), CC BY 4.0
Additional source	Creative Commons Attribution 4.0