Skip to main content
ModelScale

Gemma 4 26B A4B

Google · Open Weight · rank 89 · bench-align-v5

Self-hosted
Compare this model Model a workload
Canonical idgemma-4-26b-a4b
Overall score55.1
Context window256K tokens
Release date2026-04-02
Access typeOpen Weight
Blended $/1MUnavailable75% input / 25% output

Capability shape

Seven axes from the ranking source. A missing axis is drawn as a gap.

Capability evidence

AgenticUnavailable
CodingUnavailable
Knowledge33.3
Reasoning66.2
Multimodal & Grounded44.1
Instruction Following88.7
MathUnavailable

Runtime service evidence

Measured values with the date they were observed. Nothing is inferred from a sibling model or a provider claim.

Time to first token: latency to the first answer chunk (Artificial Analysis, via BenchLM). Reasoning models include thinking time, so values can run to tens or hundreds of seconds.

Evidence key: ObservedUnavailable

Runtime measurements
MeasurementValueObservedLast goodEvidence
Time to first tokenUnavailableUnavailableUnavailable
ThroughputUnavailableUnavailableUnavailable

Regional or per-endpoint measurements appear only when the API supplies them; none are modelled here.

Endpoint and price matrix

Every published price component, including cache reads and writes.

Price components
ComponentUSDEvidence
Input / 1M tokensUnavailable
Output / 1M tokensUnavailable
Cache read / 1M tokensUnavailable
Cache write / 1M tokensUnavailable
Blended / 1M (75% input / 25% output)Unavailable
Self-hosted listingSelf-hosted open-weight model. Public hosted pricing varies by provider.No hosted token rate was published for this model, so its per-token price is unavailable rather than zero.

Workload-aware monthly cost example

10 conversations per day × 8 messages × 22 active days, 1200 input and 400 output tokens per message, no cache. This uses the same calculator as the cost simulator, so an unavailable applicable rate makes the example unavailable too.

Modelled monthly costUnavailableThe applicable input rate is unavailable.
Modelled tokensUnavailable
Open the simulatorChange this workload

Benchmark record

18 matched benchmark rows with their published value, unit, and provenance.

Knowledge

Knowledge benchmarks
BenchmarkValueTasksDifficultyProvenance
MMLU-Pro
Massive Multitask Language Understanding Professional
82.6Multiple subjectsProfessional levelMMLU-Pro: A More Robust and Challenging Multi-Task Language Understanding Benchmark
HLE
Humanity's Last Exam
17.2Expert-level questionsFrontier expert levelHumanity's Last Exam
Artificial Analysis Intelligence Index
Artificial Analysis Intelligence Index
16.7Cross-benchmark intelligence indexDisplay-only external referenceArtificial Analysis
AA-GPQA Diamond
Artificial Analysis GPQA Diamond
79.2Graduate-level science questionsGraduate-level science reasoningArtificial Analysis GPQA Diamond Benchmark Leaderboard
AA-HLE
Artificial Analysis Humanity's Last Exam
19.3Expert-level questionsFrontier expert reasoningArtificial Analysis Humanity's Last Exam Benchmark Leaderboard
AA-Omniscience Index
Artificial Analysis Omniscience Index
-50.8Knowledge questionsBroad factual knowledgeAA-Omniscience: Knowledge and Hallucination Benchmark
AA-Omniscience Accuracy
Artificial Analysis Omniscience Accuracy
19.1Knowledge questionsBroad knowledgeArtificial Analysis model benchmarks
AA-Omniscience Hallucination Rate
Artificial Analysis Omniscience Hallucination Rate
86.4Knowledge questionsFactualityArtificial Analysis model benchmarks
HLE w/o tools
Humanity's Last Exam without tools
8.7Expert-level questionsFrontier expert levelIntroducing GPT-5.4 mini and nano

Coding

Coding benchmarks
BenchmarkValueTasksDifficultyProvenance
AA Coding Index
Artificial Analysis Coding Index
39.3Cross-benchmark coding indexDisplay-only external referenceArtificial Analysis model leaderboards

Reasoning

Reasoning benchmarks
BenchmarkValueTasksDifficultyProvenance
AA-LCR
Artificial Analysis Long Context Reasoning
65.7Long-context reasoning tasksLong-context reasoningArtificial Analysis model benchmarks
CritPt
Critical Physics Tasks
0.0Research-level physics questionsResearch-level physics reasoningCritPt Benchmark Leaderboard

Instruction Following

Instruction Following benchmarks
BenchmarkValueTasksDifficultyProvenance
AA-IFBench
Artificial Analysis IFBench
72.4Verifiable instruction constraintsInstruction precisionArtificial Analysis IFBench Benchmark Leaderboard

Agentic

Agentic benchmarks
BenchmarkValueTasksDifficultyProvenance
GDPval-AA
GDPval-AA
713Agentic real-world work tasksProfessional agentic workflowsDeepSeek-V4 Technical Report
GDPval-AA
GDPval-AA normalized
10.7Economically valuable tasksProfessional agentic workflowsArtificial Analysis model benchmarks
τ²-bench results
τ²-Bench Tool-Agent-User Evaluation
43.6Airline, retail, and telecom customer-service task setsDual-control customer-service workflowsτ²-Bench: Evaluating Conversational Agents in a Dual-Control Environment

Multimodal & Grounded

Multimodal & Grounded benchmarks
BenchmarkValueTasksDifficultyProvenance
MMMU-Pro
Massive Multi-discipline Multimodal Understanding Pro
73.8Multimodal academic reasoningFrontier multimodalMMMU-Pro: A More Robust Multi-discipline Multimodal Understanding Benchmark
AA-MMMU-Pro
Artificial Analysis MMMU-Pro
69.2Multimodal academic reasoningFrontier multimodalArtificial Analysis MMMU-Pro Benchmark Leaderboard

Lifecycle and limitations log

Lifecycle events the source associates with this model, plus what this profile does not claim.

No lifecycle event references this model.

That is not evidence the model has no lifecycle plan — only that this source published none.

What this profile does not claimValues are reproduced exactly as their sources published them, in the units those sources declared; none are converted, interpolated, or averaged across providers. Any field marked unavailable was attempted and not returned — the attempt timestamp is in each badge. Last attempted fetch for this model’s score: 2026-09-17 17:17 UTC.