LIVE INTELLIGENCE · part of the Xentir AI Index

DeepSeek-R1 (May 2025) vs Gemma 4 31B IT

Name: DeepSeek-R1 (May 2025) vs Gemma 4 31B IT — benchmark results
Creator: Epoch AI
License: https://creativecommons.org/licenses/by/4.0/

DeepSeek-R1 (May 2025) OPEN WEIGHTS · DeepSeek, released 2025-05-28
Gemma 4 31B IT OPEN WEIGHTS · Google DeepMind, released 2026-04-02

“Epoch AI, 'AI Benchmarking Hub'. Published online at epoch.ai. Retrieved from 'https://epoch.ai/benchmarks' [online resource].” · epoch.ai/benchmarks · Licensed CC BY 4.0 · data last verified 2026-09-09 · Methodology · Data Sources

Gemma 4 31B IT leads on more of the 3 benchmarks where both are measured — DeepSeek-R1 (May 2025) on 1, Gemma 4 31B IT on 2. The remaining 6 of 9 tracked benchmarks cannot be compared: 3 have a result for only one of the two models, 3 for neither. Nothing on this page is estimated to fill those in.

Head-to-head

Measured on 3 of 9 tracked benchmarks for both models — DeepSeek-R1 (May 2025) leads on 1, Gemma 4 31B IT leads on 2.

Benchmark	DeepSeek-R1 (May 2025)	Gemma 4 31B IT	Leads
Epoch Capabilities Index (ECI)	141.3	142.7	Gemma 4 31B IT
SWE-bench Verified	Not measured	Not measured	Not measured on both
GPQA Diamond	0.763	0.758 (minimal)	DeepSeek-R1 (May 2025)
MATH Level 5	0.966	Not measured	Comparison unavailable — Gemma 4 31B IT has no published result
OTIS Mock AIME 2024-2025	0.664	0.733 (minimal)	Gemma 4 31B IT
FrontierMath	Not measured	Not measured	Not measured on both
FrontierMath Tier 4	Not measured	Not measured	Not measured on both
SimpleQA Verified	Not measured	0.104	Comparison unavailable — DeepSeek-R1 (May 2025) has no published result
Chess Puzzles	Not measured	0.050 (minimal)	Comparison unavailable — DeepSeek-R1 (May 2025) has no published result

Benchmark results measure specific tasks and should not be treated as a universal measure of model quality. A model with no row here for a benchmark has not been measured on it — that is left blank, never estimated.

What remains uncertain

DeepSeek-R1 (May 2025) has no published result on Chess Puzzles, SimpleQA Verified.
Gemma 4 31B IT has no published result on MATH Level 5.
3 tracked benchmarks have no published result for either model.

A missing result means Xentir has not found one published by the source it cites, not that the model performs badly. See Data Sources.

Full profiles

DeepSeek-R1 (May 2025) → · Gemma 4 31B IT → · the full AI Model Rankings table