Compare AI models

Specifications
Parameters
236B1B/4B/12B/27B
Context window
128k—
API pricing
Cheapest input
—$0.04DeepInfra
Cheapest output
—$0.08DeepInfra
—$15.00DeepInfra · bf16
Benchmarks
HumanEval
84.76%87.8%
MATH
71.02%89%
Overview
CompanyDeepSeekGoogle
Release dateJun 28 2024Mar 12 2025
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Gemma 3 leads DeepSeek V2-0628 on 2 of the 2 benchmarks they both report (HumanEval, MATH). DeepSeek V2-0628 shipped 257 days before Gemma 3, so benchmark comparisons should account for the intervening progress.

DeepSeek V2-0628 has 236B parameters, while Gemma 3 has 1B/4B/12B/27B.

On HumanEval, Gemma 3 leads at 87.8% vs DeepSeek V2-0628 at 84.76%. On MATH, Gemma 3 leads at 89% vs DeepSeek V2-0628 at 71.02%.