← All models
Compare

Gemini 3.1 Pro vs GPT-5.4

Add, remove, or swap models to compare them side by side.

Gemini 3.1 ProGPT-5.4
AttributeGemini 3.1 ProGoogle DeepMind · 3 in the newsGPT-5.4OpenAI · 9 in the news
Scores
Intelligence (ECI)155157
Coding6277
Math7288
Reasoning & Knowledge9478
Agentic & Tools8172
Specifications
DeveloperGoogle DeepMindOpenAI
FamilyGeminiGPT
ReleasedFeb 19, 2026Mar 5, 2026
Parameters
AvailabilityAPI accessAPI access
Context window1.1M
Price — $/M input$2.50
Price — $/M output$15.00
Inputstext, image, file
Outputstext
Benchmarks
AIME 2024/202596%98%
APEX34%36%
ARC-AGI98%94%
ARC-AGI-277%74%
FrontierMath37%48%
FrontierMath Tier 417%27%
GPQA Diamond94%93%
GSO (code optimization)23%31%
Humanity's Last Exam46%36%
METR task horizon6.4 h5.7 h
SimpleBench80%
SimpleQA Verified77%45%
SWE-bench Verified76%77%
Terminal-Bench80%82%
WebDev Arena14611457
WeirdML72%78%
SciCode57%
τ²-bench87%

Use-case scores are 0–100 percentile composites across each area’s benchmarks, ranked against every model from the past year. Highlighted cells lead each row. Open a model for the full picture.

Benchmarks & model data from Epoch AI (CC BY); pricing & specs from OpenRouter. ECI = Epoch Capabilities Index.