The AI arena is free today

Open Superagent

MAI-Thinking-1 vs Qwen3.5-35B-A3B

MAI-Thinking-1 and Qwen3.5-35B-A3B are closely matched at 33.0 and 30.9 on the LLM Stats Score.

Microsoft · Alibaba Cloud / Qwen Team · Updated for 2026

Which is better?

MAI-Thinking-1 and Qwen3.5-35B-A3B are closely matched on the overall LLM Stats Score at 33.0 and 30.9.

In the 9 individual benchmarks reported for both models, Qwen3.5-35B-A3B wins 5; this is a narrower head-to-head signal than the composite indexes.

Based on current LLM Stats indexes, shared benchmarks, pricing, and model metadata for 2026.

Choose MAI-Thinking-1

  • you want the most recent training data — it shipped Jun 2026

Choose Qwen3.5-35B-A3B

  • you value its reported benchmark strengths — it wins 5 of 9 exact shared results
  • you need open weights you can self-host or fine-tune

At a glance

The differences that matter most.

Core performance indexes
33.0
#101
30.9
#114
33.8
#92
31.9
#104
19.4
#104
11.6
#152
12.8
#97
12.6
#101
Cost, coverage & limits
Benchmark wins
4 of 9
5 of 9
Input price
— / M
$0.14 / M
Output price
— / M
$1.00 / M
Context window
262,144

Capability indexes

Additional strengths measured across groups of related public benchmarks

5 shared
Index
MAI-Thinking-1
Qwen3.5-35B-A3B
33.7#53
32.7#55
10.7#119
11.6#109
20.4#30
20.4#29
29.2#29
31.1#20
26.7#31
31.1#19
Conservative TrueSkill rating · higher is betterHow scores work

Individual benchmarks

23 reported for MAI-Thinking-1 · 81 for Qwen3.5-35B-A3B

9 shared

MAI-Thinking-1 outperforms in 4 benchmarks (LiveCodeBench v6, LongBench v2, SWE-Bench Verified, Terminal-Bench 2.0), while Qwen3.5-35B-A3B is better at 4 benchmarks (IFBench, MedXpertQA, MMLU-Pro, Multi-Challenge).

Both models are evenly matched across the benchmarks.

Sat Sep 12 2026 • llm-stats.com

Human preference

Blind head-to-head votes and playground preference scores

Model Size

Parameter count comparison

965.0B diff

MAI-Thinking-1 has 965.0B more parameters than Qwen3.5-35B-A3B, making it 2757.1% larger.

Microsoft
MAI-Thinking-1
1.0Tparameters
Alibaba Cloud / Qwen Team
Qwen3.5-35B-A3B
35.0Bparameters
1000.0B
MAI-Thinking-1
35.0B
Qwen3.5-35B-A3B

Context Window

Maximum input and output token capacity

Only Qwen3.5-35B-A3B specifies input context (262,144 tokens). Only Qwen3.5-35B-A3B specifies output context (262,144 tokens).

Microsoft
MAI-Thinking-1
Input- tokens
Output- tokens
Alibaba Cloud / Qwen Team
Qwen3.5-35B-A3B
Input262,144 tokens
Output262,144 tokens
Sat Sep 12 2026 • llm-stats.com

Input capabilities

Documented input modalities across available providers

Qwen3.5-35B-A3B supports multimodal inputs, whereas MAI-Thinking-1 does not.

Qwen3.5-35B-A3B can handle both text and other forms of data like images, making it suitable for multimodal applications.

MAI-Thinking-1

Text
Images
Audio
Video

Qwen3.5-35B-A3B

Text
Images
Audio
Video

License

Usage and distribution terms

MAI-Thinking-1 is licensed under a proprietary license, while Qwen3.5-35B-A3B uses Apache 2.0.

License differences may affect how you can use these models in commercial or open-source projects.

MAI-Thinking-1

Proprietary

Closed source

Qwen3.5-35B-A3B

Apache 2.0

Open weights

Release Timeline

When each model was launched

MAI-Thinking-1 was released on 2026-06-02, while Qwen3.5-35B-A3B was released on 2026-02-24.

MAI-Thinking-1 is 3 months newer than Qwen3.5-35B-A3B.

MAI-Thinking-1

Jun 2, 2026

3 months ago

3mo newer
Qwen3.5-35B-A3B

Feb 24, 2026

6 months ago

Knowledge Cutoff

When training data ends

Neither model specifies a knowledge cutoff date.

Unable to compare the recency of their training data.

No cutoff dates available

Outputs Comparison

Notice missing or incorrect data?Start an Issue discussion

Judge for yourself.

Run your own prompts against MAI-Thinking-1 and Qwen3.5-35B-A3B side-by-side, then vote on the output you prefer.

MAI-Thinking-1
✓ Preferred
Qwen3.5-35B-A3B
Open in Playground

FAQ

Common questions about MAI-Thinking-1 vs Qwen3.5-35B-A3B.

Which is better, MAI-Thinking-1 or Qwen3.5-35B-A3B?

MAI-Thinking-1 and Qwen3.5-35B-A3B are closely matched on the LLM Stats Score at 33.0 and 30.9. MAI-Thinking-1 is made by Microsoft and Qwen3.5-35B-A3B is made by Alibaba Cloud / Qwen Team. The best choice depends on your use case — compare their capability indexes, individual benchmarks, pricing, and limits above.

How does MAI-Thinking-1 compare to Qwen3.5-35B-A3B in benchmarks?

MAI-Thinking-1 scores LongFact: 98.0%, AIME 2025: 97.0%, AIME 2026: 94.5%, GraphWalks: 90.0%, AIR-Bench: 88.0%. Qwen3.5-35B-A3B scores CountBench: 97.8%, VLMsAreBlind: 97.0%, MMLU-Redux: 93.3%, V*: 92.7%, AI2D: 92.6%.

What are the context window sizes for MAI-Thinking-1 and Qwen3.5-35B-A3B?

MAI-Thinking-1 supports an unknown number of tokens and Qwen3.5-35B-A3B supports 262K tokens. A larger context window lets you process longer documents, conversations, or codebases in a single request.

What are the main differences between MAI-Thinking-1 and Qwen3.5-35B-A3B?

Key differences include LLM Stats Score (33.0 vs 30.9), multimodal support (no vs yes), licensing (Proprietary vs Apache 2.0). See the full comparison above for benchmark-by-benchmark results.

Who makes MAI-Thinking-1 and Qwen3.5-35B-A3B?

MAI-Thinking-1 is developed by Microsoft and Qwen3.5-35B-A3B is developed by Alibaba Cloud / Qwen Team.