DataLearner logo

Qwen3.6-35B-A3BvsGemma 4 26B A4B

Across 4 shared benchmarks, Qwen3.6-35B-A3B leads overall: Qwen3.6-35B-A3B wins 4, Gemma 4 26B A4B wins 0, with 0 ties and an average score difference of +3.62.

阿里巴巴
Qwen3.6-35B-A3B

阿里巴巴 · 2026-04-16 · Reasoning model

DeepMind
Gemma 4 26B A4B

DeepMind · 2026-04-02 · Chat model

Qwen3.6-35B-A3B4 wins(100%)(0%)0 winsGemma 4 26B A4B

Benchmark scores

Grouped by capability, sorted by largest gap within each. 4 shared benchmarks.

General Knowledge

Qwen3.6-35B-A3B 2/2
BenchmarkQwen3.6-35B-A3BGemma 4 26B A4BDiff
HLE21.40123 / 181Thinking (No Tools)17.20140 / 181Thinking (With Tools + Internet)+4.20
MMLU Pro85.2025 / 133Thinking (No Tools)82.6050 / 133Thinking (No Tools)+2.60

Coding and Software Engineer

Qwen3.6-35B-A3B 1/1
BenchmarkQwen3.6-35B-A3BGemma 4 26B A4BDiff
LiveCodeBench80.4030 / 126Thinking (No Tools)77.1036 / 126Thinking (No Tools)+3.30

Math and Reasoning

Qwen3.6-35B-A3B 1/1
BenchmarkQwen3.6-35B-A3BGemma 4 26B A4BDiff
AIME 202692.7010 / 19Thinking (No Tools)88.3017 / 19Thinking (No Tools)+4.40

Specs

FieldQwen3.6-35B-A3BGemma 4 26B A4B
Publisher阿里巴巴DeepMind
Release date2026-04-162026-04-02
Model typeReasoning modelChat model
ArchitectureMoEMoE
Parameters35B25.2B
Context length200K256K
Max output80K32K

Summary

  • Qwen3.6-35B-A3Bleads in:General Knowledge (2/2), Coding and Software Engineer (1/1), Math and Reasoning (1/1)

On average across the 4 shared benchmarks, Qwen3.6-35B-A3B scores 3.62 higher.

Largest single-benchmark gap: AIME 2026 — Qwen3.6-35B-A3B 92.70 vs Gemma 4 26B A4B 88.30 (+4.40).

Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.