Skip to main content

Gemma 4 31B vs Qwen3 235B-A22B Thinking 2507

SpecGemma 4 31BQwen3 235B-A22B Thinking 2507
Respan Index18.7 (#139)18.5 (#141)
Blended price per 1M-$1.21
Context window256K262K
Input / output per 1M$0 / $0$0.45 / $3.5
Max output--
Inputtext, visiontext
Released2026-04-022025-07-25

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.

CategoryGemma 4 31BQwen3 235B-A22B Thinking 2507
Coding7.8 Gemma 4 31B5.2 Qwen3 235B-A22B Thinking 2507
Agents12.5* Gemma 4 31B4.2 Qwen3 235B-A22B Thinking 2507
Reasoning19.7 Gemma 4 31B22.1 Qwen3 235B-A22B Thinking 2507
Knowledge & language11.0 Gemma 4 31B38.2 Qwen3 235B-A22B Thinking 2507
Human preference50.5 Gemma 4 31B45.5 Qwen3 235B-A22B Thinking 2507

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

Benchmarks both report

Gemma 4 31B leads on 4 of the 7 benchmarks both report. Higher is better. Each score links to where it was published.

BenchmarkGemma 4 31BQwen3 235B-A22B Thinking 2507
LMArena Elo14521400
Chess Puzzles (Epoch AI)5%12%
GPQA Diamond84.3%80.1%
OTIS Mock AIME 2024-2025 (Epoch AI)73.3%86.7%
SimpleQA Verified10.4%40.4%
WeirdML (Håvard Tveit Ihle)52.3%41%
MMLU-Pro85.2%84.4%

FAQ

Is Gemma 4 31B better than Qwen3 235B-A22B Thinking 2507?

Gemma 4 31B scores 18.7 on the Respan Index (#139), ahead of Qwen3 235B-A22B Thinking 2507 at 18.5 (#141). Gemma 4 31B leads on 4 of the 7 benchmarks both report.

Which has the larger context window?

Gemma 4 31B accepts 256K tokens and Qwen3 235B-A22B Thinking 2507 262K tokens.

Compare other models

vs

Built for AI agents.
Break less.
Ship more.