DataLearner logo

Qwen 3.6 Plus PreviewvsQwen3.5-397B-A17B

Across 11 shared benchmarks, Qwen 3.6 Plus Preview leads overall: Qwen 3.6 Plus Preview wins 10, Qwen3.5-397B-A17B wins 1, with 0 ties and an average score difference of +3.04.

阿里巴巴
Qwen 3.6 Plus Preview

阿里巴巴 · 2026-03-31 · Chat model

阿里巴巴
Qwen3.5-397B-A17B

阿里巴巴 · 2026-02-16 · Multimodal model

Qwen 3.6 Plus Preview10 wins(91%)(9%)1 winQwen3.5-397B-A17B

Benchmark scores

Grouped by capability, sorted by largest gap within each. 11 shared benchmarks.

Coding and Software Engineer

Qwen 3.6 Plus Preview 4/4
BenchmarkQwen 3.6 Plus PreviewQwen3.5-397B-A17BDiff
SWE-Bench Pro - Public56.6025 / 62Thinking (No Tools)50.9045 / 62Thinking (No Tools)+5.70
SWE-bench Multilingual73.8014 / 30Thinking (No Tools)69.3025 / 30Thinking (No Tools)+4.50
LiveCodeBench87.1019 / 250Thinking (No Tools)83.6037 / 250Thinking (No Tools)+3.50
SWE-bench Verified78.8021 / 116Thinking (With Tools)76.4033 / 116Thinking (With Tools)+2.40

AI Agent - Tool Usage

Qwen 3.6 Plus Preview 2/2
BenchmarkQwen 3.6 Plus PreviewQwen3.5-397B-A17BDiff
Terminal Bench 2.061.6016 / 48Thinking (With Tools)52.5031 / 48Thinking (With Tools)+9.10
Tool Decathlon39.807 / 10Thinking (With Tools)38.308 / 10Thinking (With Tools)+1.50

General Knowledge

Qwen 3.6 Plus Preview 2/2
BenchmarkQwen 3.6 Plus PreviewQwen3.5-397B-A17BDiff
MMLU Pro88.506 / 176Thinking (No Tools)87.8011 / 176Thinking (No Tools)+0.70
C-Eval93.302 / 48Thinking (No Tools)933 / 48Thinking (No Tools)+0.30

Math and Reasoning

Qwen 3.6 Plus Preview 2/2
BenchmarkQwen 3.6 Plus PreviewQwen3.5-397B-A17BDiff
AIME 202695.3012 / 29Thinking (No Tools)91.3022 / 29Thinking (No Tools)+4
IMO-AnswerBench83.8014 / 24Thinking (No Tools)80.9019 / 24Thinking (No Tools)+2.90

Long Context

Qwen3.5-397B-A17B 1/1
BenchmarkQwen 3.6 Plus PreviewQwen3.5-397B-A17BDiff
LongBench v2624 / 14Normal (No Tools)63.203 / 14Normal (No Tools)-1.20

Specs

FieldQwen 3.6 Plus PreviewQwen3.5-397B-A17B
Publisher阿里巴巴阿里巴巴
Release date2026-03-312026-02-16
Model typeChat modelMultimodal model
ArchitectureDenseMoE
ParametersNot available397B
Context length1M256K
Max output64KNot available

API pricing

Prices use DataLearner records when available; missing fields are not inferred.

ItemQwen 3.6 Plus PreviewQwen3.5-397B-A17B
Text input$0.5 / 1M tokens$0.5 / 1M tokens
Text output$3 / 1M tokens$3 / 1M tokens
Cache read$0.05 / 1M tokens$0.05 / 1M tokens
Cache write$0.625 / 1M tokens$0.625 / 1M tokens

Summary

  • Qwen 3.6 Plus Previewleads in:Coding and Software Engineer (4/4), AI Agent - Tool Usage (2/2), General Knowledge (2/2), Math and Reasoning (2/2)
  • Qwen3.5-397B-A17Bleads in:Long Context (1/1)

On average across the 11 shared benchmarks, Qwen 3.6 Plus Preview scores 3.04 higher.

Largest single-benchmark gap: Terminal Bench 2.0 — Qwen 3.6 Plus Preview 61.60 vs Qwen3.5-397B-A17B 52.50 (+9.10).

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