← All models
Compare

gpt-oss-20b vs Grok 4.5

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

gpt-oss-20bGrok 4.5
Attributegpt-oss-20bOpenAI · 1 in the newsGrok 4.5xAI · 2 in the news
Scores
Intelligence (ECI)154
Coding1989
Math4784
Reasoning & Knowledge1168
Agentic & Tools1174
Specifications
DeveloperOpenAIxAI
FamilyGPTGrok
ReleasedAug 5, 2025Jul 8, 2026
Parameters20.9B
AvailabilityOpen weights (unrestricted)API access
Context window131K500K
Price — $/M input$0.03$2.00
Price — $/M output$0.14$6.00
Inputstexttext, image, file
Outputstexttext
Benchmarks
AIME 2024/202589%98%
GPQA Diamond69%93%
Humanity's Last Exam10%40%
LiveCodeBench78%
MMLU-Pro75%
SciCode34%54%
Terminal-Bench3%
WeirdML41%46%
τ²-bench60%
APEX34%
ARC-AGI87%
ARC-AGI-253%
SimpleBench70%
SimpleQA Verified54%
WebDev Arena1551

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.