Compare AI models

Specifications
Parameters
671B5.1B/8B/26B/31B
Context window
128k—
API pricing
Cheapest input
—$0.04Darkbloom
Cheapest output
—$0.10Modal
—$15.50Darkbloom
Benchmarks
LiveCodeBench
65.9%80%
GPQA Diamond
71.5%84.3%
MMLU-Pro
84%85.2%
Overview
CompanyDeepSeekGoogle
Release dateJan 20 2025Apr 2 2026
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Gemma 4 leads DeepSeek R1 on 3 of the 3 benchmarks they both report (LiveCodeBench, GPQA Diamond, MMLU-Pro). DeepSeek R1 shipped 437 days before Gemma 4, so benchmark comparisons should account for the intervening progress.

DeepSeek R1 has 671B parameters, while Gemma 4 has 5.1B/8B/26B/31B.

On LiveCodeBench, Gemma 4 leads at 80% vs DeepSeek R1 at 65.9%. On GPQA Diamond, Gemma 4 leads at 84.3% vs DeepSeek R1 at 71.5%. On MMLU-Pro, Gemma 4 leads at 85.2% vs DeepSeek R1 at 84%.