Gemma 4 26B A4B Benchmark Details

Gemma 4 26B A4B currently shows benchmark results led by MMLU-Pro (51 / 175, score 82.60), LiveCodeBench (37 / 125, score 77.10), GPQA Diamond (114 / 253, score 82.30). This page also compares it with 2 competitor models and 1 predecessor or same-series models, including performance and pricing views when available.

Benchmark Results

Gemma 4 26B A4B

Benchmark Results

Knowledge Exams

3 evaluations

Benchmark / mode

Score

Rank/total

MMLU-Pro

Thinking Mode

82.60

51 / 175

HLE

Thinking Mode

8.70

201 / 235

HLE

Thinking ModeToolsInternet

17.20

174 / 235

Scientific Reasoning

1 evaluations

Benchmark / mode

Score

Rank/total

GPQA Diamond

Thinking Mode

82.30

114 / 253

Algorithmic Coding

2 evaluations

Benchmark / mode

Score

Rank/total

CodeForces

Thinking Mode

1718

18 / 21

LiveCodeBench

Thinking Mode

77.10

37 / 125

Writing

1 evaluations

Benchmark / mode

Score

Rank/total

Creative Writing

Standard Mode

1304.60

77 / 110

Service Workflows

2 evaluations

Benchmark / mode

Score

Rank/total

τ²-Bench

Thinking ModeTools

68.20

27 / 44

τ³-Banking

Thinking ModeTools

126 / 167

Agentic Development

2 evaluations

Benchmark / mode

Score

Rank/total

Terminal Bench Hard

Standard ModeTools

128 / 244

Terminal Bench Hard

Thinking ModeTools

13.60

173 / 244

Mathematics

1 evaluations

Benchmark / mode

Score

Rank/total

AIME 2026

Thinking Mode

88.30

27 / 30

Visual Understanding

4 evaluations

Benchmark / mode

Score

Rank/total

MathVision

Thinking Mode

82.40

10 / 12

MMMU-Pro

Standard Mode

66.70

148 / 229

MMMU-Pro

unknown

73.80

104 / 229

MMMU-Pro

Thinking Mode

69.20

134 / 229

Long Retrieval

1 evaluations

Benchmark / mode

Score

Rank/total

GDM-MRCR v2 (8-needle, 128K)

Thinking Mode

44.10

6 / 8

Tool Orchestration

1 evaluations

Benchmark / mode

Score

Rank/total

PinchBench v2

unknown

56.37

34 / 45

Compare with other models

Competitor Comparison

Benchmark scores for Gemma 4 26B A4B compared against top models in its class

Gemma 4 26B A4BQwen3.6-35B-A3B GLM-4.7-Flash

Benchmark categories:

The chart shows each model’s highest score per benchmark within the current filter. Out-of-100 benchmarks use raw heights; out-of-range benchmarks are scaled within that benchmark while labels keep the original scores.

10 benchmarks with comparable scores. Each model shows its best score; mode label is displayed below.

Benchmark	Gemma 4 26B A4BCurrent	Qwen3.6-35B-A3B	GLM-4.7-Flash
HLE Accuracy Knowledge Exams	17.20Thinking Enabled ｜ Tools	21.40Thinking Enabled	14.40Thinking Enabled
MMLU-Pro Accuracy Knowledge Exams	82.60Thinking Enabled	85.20Thinking Enabled	--
GPQA Diamond Accuracy Scientific Reasoning	82.30Thinking Enabled	84.85Standard Mode	75.20Thinking Enabled
LiveCodeBench Pass @K Algorithmic Coding	77.10Thinking Enabled	80.40Thinking Enabled	--
Creative Writing Elo、大模型评判两两对战 Writing	1304.60Standard Mode	--	1124.70Standard Mode
τ²-Bench Accuracy Service Workflows	68.20Thinking Enabled ｜ Tools	--	79.50Thinking Enabled ｜ Tools
τ³-Banking Score Service Workflows	12.00Thinking Enabled ｜ Tools	9.30Thinking Enabled ｜ Tools	--
Terminal Bench Hard Accuracy Agentic Development	25.00Standard Mode ｜ Tools	34.80Thinking Enabled ｜ Tools	32.00Thinking Enabled ｜ Tools
AIME 2026 Accuracy Mathematics	88.30Thinking Enabled	92.70Thinking Enabled	--
MMMU-Pro Accuracy Visual Understanding	73.80Thinking Level · High	75.00Thinking Enabled	--

Standard API Pricing: Gemma 4 26B A4B vs. Peer Models

Shows standard text input and output pricing side by side for each model. If extended-context pricing exists, the chart keeps the base rate and explains the threshold below.

Source: DataLearnerAI. Standard text prices shown here use the default supplier. · CNY / 1M tokens

Model	Supplier	Standard input	Standard output	Base price applies to
GLM-4.7-Flash	智谱AI	¥0 / 1M tokens	¥0 / 1M tokens	—

Version History

How each version of the Gemma 4 26B A4B series stacks up on benchmark tests

Gemma 4 26B A4BGemma 3 - 27B (IT)

Benchmark categories:

4 benchmarks with comparable scores. Each model shows its best score; mode label is displayed below.· Click a row to view its trend chart.

Benchmark	Gemma 4 26B A4BCurrent	Gemma 3 - 27B (IT)
MMLU-Pro Accuracy Knowledge Exams	82.60Thinking Enabled	67.50Standard Mode
GPQA Diamond Accuracy Scientific Reasoning	82.30Thinking Enabled	42.40Standard Mode
LiveCodeBench Pass @K Algorithmic Coding	77.10Thinking Enabled	29.70Standard Mode
Creative Writing Elo、大模型评判两两对战 Writing	1304.60Standard Mode	1265.70Standard Mode

Single-Benchmark Version Trend

Viewing: MMLU-Pro · Knowledge Exams

Benchmark

NormalNormal + ToolsThinkingThinking + ToolsDeepDeep + Tools

X-axis shows model and release date, Y-axis shows score; solid lines connect the same mode across versions, while dotted guides align modes within the same generation.

Standard API Pricing Across the Gemma 4 26B A4B Series

Shows standard text input and output pricing side by side for each model. If extended-context pricing exists, the chart keeps the base rate and explains the threshold below.

Source: DataLearnerAI. Standard text prices shown here use the default supplier. · USD / 1M tokens

Model	Supplier	Standard input	Standard output	Base price applies to
Gemma 3 - 27B (IT)	DeepInfra	$0.09 / 1M tokens	$0.16 / 1M tokens	—