Skip to main content

DeepSeek R1 vs Gemma 4 31B

SpecDeepSeek R1Gemma 4 31B
Respan Index18.8 (#137)18.7 (#139)
Blended price per 1M$1.15-
Context window128K256K
Input / output per 1M$0.7 / $2.5$0 / $0
Max output33K-
Inputtexttext, vision
Released2025-01-202026-04-02

Respan Index by category

Each category is scored 0 to 100, and how far ahead the winner is. Leads under a point are close to even.

CategoryDeepSeek R1Gemma 4 31B
Coding3.6 DeepSeek R17.8 Gemma 4 31B
Agents16.4 DeepSeek R112.5* Gemma 4 31B
Reasoning21.2 DeepSeek R119.7 Gemma 4 31B
Knowledge & language17.5 DeepSeek R111.0 Gemma 4 31B
Human preference45.2 DeepSeek R150.5 Gemma 4 31B

* Estimated: no published scores in that category. How the Respan Index works

Benchmarks both report

Gemma 4 31B leads on 5 of the 5 benchmarks both report. Higher is better. Each score links to where it was published.

BenchmarkDeepSeek R1Gemma 4 31B
LMArena Elo13981452
GPQA Diamond71.5%84.3%
OTIS Mock AIME 2024-2025 (Epoch AI)53.3%73.3%
WeirdML (Håvard Tveit Ihle)36.5%52.3%
MMLU-Pro84%85.2%

FAQ

Is DeepSeek R1 better than Gemma 4 31B?

DeepSeek R1 scores 18.8 on the Respan Index (#137), ahead of Gemma 4 31B at 18.7 (#139). Gemma 4 31B leads on 5 of the 5 benchmarks both report.

Which has the larger context window?

DeepSeek R1 accepts 128K tokens and Gemma 4 31B 256K tokens.

Compare other models

vs

Built for AI agents.
Break less.
Ship more.