Compare AI models

Specifications
Parameters
671B117B
Context window
128k128k
API pricing
Cheapest input
—$0.03AkashML
Cheapest output
—$0.15Near AI
—$13.50Venice · int4
Benchmarks
SWE-Bench Verified
49.2%62.4%
AIME 2024
79.8%95.8%
GPQA Diamond
71.5%80.1%
MMLU
90.8%90%
Overview
CompanyDeepSeekOpenAI
Release dateJan 20 2025Aug 5 2025
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

gpt-oss-120b leads DeepSeek R1 on 3 of the 4 benchmarks they both report (SWE-Bench Verified, AIME 2024, GPQA Diamond, MMLU). DeepSeek R1 shipped 197 days before gpt-oss-120b, so benchmark comparisons should account for the intervening progress.

DeepSeek R1 has 671B parameters, while gpt-oss-120b has 117B. Context windows are 128k (DeepSeek R1) vs 128k (gpt-oss-120b).

On SWE-Bench Verified, gpt-oss-120b leads at 62.4% vs DeepSeek R1 at 49.2%. On AIME 2024, gpt-oss-120b leads at 95.8% vs DeepSeek R1 at 79.8%. On GPQA Diamond, gpt-oss-120b leads at 80.1% vs DeepSeek R1 at 71.5%. On MMLU, DeepSeek R1 leads at 90.8% vs gpt-oss-120b at 90%.