# Gemma 4 26B A4B Benchmark Scores & Performance

> Gemma 4 26B A4B by Google scores 46.88/100 overall, ranking #108 out of 638 AI models.

Canonical page: https://benchlm.ai/models/gemma-4-26b-a4b

Last updated: October 1, 2026

## Model Details

| Property | Value |
|----------|-------|
| Creator | Google |
| Source Type | Open Weight |
| Reasoning Type | Reasoning |
| Context Window | 256K |
| Overall Score | 46.88/100 |
| Overall Rank | #108 of 638 |

## Family & Coverage

- Family: Gemma 4
- Variant: 26b-a4b
- Benchmarks covered: 19 of 508
- Sibling models: [Gemma 4 31B](/models/gemma-4-31b), [Gemma 4 12B](/models/gemma-4-12b), [Gemma 4 E4B](/models/gemma-4-e4b), [Gemma 4 E2B](/models/gemma-4-e2b)
- Coverage note: BenchLM currently has partial benchmark coverage for this model, so the overall score is conservative.

## Agentic Benchmarks

| Benchmark | Score | Source |
|-----------|-------|--------|
| [τ²-bench results](/benchmarks/tau2-bench) | 43.6% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/gemma-4-26b-a4b) |
| [GDPval-AA](/benchmarks/gdpvalaanormalized) | 2.6% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/gemma-4-26b-a4b) |
| [GDPval-AA](/benchmarks/gdpvalaa) | 713 | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/gemma-4-26b-a4b) |

## Coding Benchmarks

| Benchmark | Score | Source |
|-----------|-------|--------|
| [AA Coding Index](/benchmarks/aacodingindex) | 39.3% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/gemma-4-26b-a4b) |
| [AA-SciCode](/benchmarks/aascicode) | 40.0% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/gemma-4-26b-a4b) |

## Multimodal & Grounded Benchmarks

| Benchmark | Score | Source |
|-----------|-------|--------|
| [MMMU-Pro](/benchmarks/mmmu-pro) | 73.8% | [Gemma 4 26B A4B model card](https://huggingface.co/google/gemma-4-26B-A4B) |
| [AA-MMMU-Pro](/benchmarks/aammmupro) | 69.2% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/gemma-4-26b-a4b) |

## Reasoning Benchmarks

| Benchmark | Score | Source |
|-----------|-------|--------|
| [AA-LCR](/benchmarks/lcr) | 65.7% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/gemma-4-26b-a4b) |
| [CritPt](/benchmarks/critpt) | 0.0% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/gemma-4-26b-a4b) |

## Knowledge Benchmarks

| Benchmark | Score | Source |
|-----------|-------|--------|
| [MMLU-Pro](/benchmarks/mmlu-pro) | 82.6% | [Gemma 4 26B A4B model card](https://huggingface.co/google/gemma-4-26B-A4B) |
| [HLE](/benchmarks/hle) | 17.2% | [Gemma 4 26B A4B model card](https://huggingface.co/google/gemma-4-26B-A4B) |
| [HLE w/o tools](/benchmarks/hlenotools) | 8.7% | [Gemma 4 26B A4B model card](https://huggingface.co/google/gemma-4-26B-A4B) |
| [Artificial Analysis Intelligence Index](/benchmarks/artificialanalysis) | 16.7% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/gemma-4-26b-a4b) |
| [AA-GPQA Diamond](/benchmarks/aagpqadiamond) | 79.2% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/gemma-4-26b-a4b) |
| [AA-HLE](/benchmarks/aahle) | 19.3% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/gemma-4-26b-a4b) |
| [AA-Omniscience Index](/benchmarks/aaomniscienceindex) | -50.8% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/gemma-4-26b-a4b) |
| [AA-Omniscience Accuracy](/benchmarks/omniscienceaccuracy) | 19.1% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/gemma-4-26b-a4b) |
| [AA-Omniscience Hallucination Rate](/benchmarks/omnisciencehallucinationrate) | 86.4% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/gemma-4-26b-a4b) |

## Instruction Following Benchmarks

| Benchmark | Score | Source |
|-----------|-------|--------|
| [AA-IFBench](/benchmarks/aaifbench) | 72.4% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/gemma-4-26b-a4b) |

## Other Google Models

- [Gemini 4 Argon](/models/gemini-4-argon) - Score: 77.18
- [Gemini 3.8 Flash](/models/gemini-3-8-flash) - Score: 73.92
- [Gemini 3.7 Flash](/models/gemini-3-7-flash) - Score: 67.94
- [Gemini 3.6 Flash](/models/gemini-3-6-flash) - Score: 65.15
- [Gemini 3.1 Pro](/models/gemini-3-1-pro) - Score: 65.05
- [Gemini 3.5 Flash](/models/gemini-3-5-flash) - Score: 63.93
- [Gemini 3 Pro](/models/gemini-3-pro) - Score: 61.55
- [Gemini 3 Flash](/models/gemini-3-flash) - Score: 56.29
- [Gemini 3.5 Flash-Lite](/models/gemini-3-5-flash-lite) - Score: 51.53
- [Gemini 2.5 Pro](/models/gemini-2-5-pro) - Score: 50.96
- [Gemini 3.1 Flash-Lite](/models/gemini-3-1-flash-lite) - Score: 49.49
- [Gemini 2.5 Flash](/models/gemini-2-5-flash) - Score: 43.73
- [Gemma 4 31B](/models/gemma-4-31b) - Score: 41.07
- [Gemma 4 12B](/models/gemma-4-12b) - Score: 32.08
- [Gemma 4 E4B](/models/gemma-4-e4b) - Score: 31.47
- [Gemma 4 E2B](/models/gemma-4-e2b) - Score: 30.34
- [Gemma 3 27B](/models/gemma-3-27b) - Score: 29.46
- [Gemini 1.5 Pro](/models/gemini-1-5-pro) - Score: 28.15
- [Gemini 1.0 Pro](/models/gemini-1-0-pro) - Score: 13.93
- [Gemini 3 Pro Deep Think](/models/gemini-3-pro-deep-think) - Score: not computed
- [Gemini 3.8 Flash TTS](/models/gemini-3-8-flash-tts) - Score: not computed
- [Gemini 3.8 Flash-Lite TTS](/models/gemini-3-8-flash-lite-tts) - Score: not computed
- [Gemini 3.8 Live](/models/gemini-3-8-live) - Score: not computed
- [Gemini 3.8 Live Extended Thinking](/models/gemini-3-8-live-extended-thinking) - Score: not computed
- [Gemini 3.8 Flash Cyber](/models/gemini-3-8-flash-cyber) - Score: not computed
- [Gemini 3.5 Transcribe](/models/gemini-3-5-transcribe) - Score: not computed
- [Gemini 3.5 Flash Cyber](/models/gemini-3-5-flash-cyber) - Score: not computed
- [Gemini 3.1 Flash TTS Preview](/models/gemini-3-1-flash-tts-preview) - Score: not computed
- [Gemini 2.5 Flash Native Audio Preview (12-2025)](/models/gemini-2-5-flash-native-audio-preview-12-2025) - Score: not computed
- [Gemini 2.5 Flash TTS Preview](/models/gemini-2-5-flash-tts-preview) - Score: not computed
- [Gemini 2.5 Pro TTS Preview](/models/gemini-2-5-pro-tts-preview) - Score: not computed
- [Gemini 3.1 Flash Live Preview](/models/gemini-3-1-flash-live-preview) - Score: not computed