Model comparison

MAI-Thinking-1 vs Qwen3.8 Max

Updated August 3, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.

MAI-Thinking-1

Microsoft

50.1/100

Estimated · Public rank #117

90% interval 40.2–60.0

Qwen3.8 Max

Alibaba

65.4/100

Estimated · Public rank #31

90% interval 55.5–75.3

Qwen3.8 Max has the higher public score estimate, 65.4 versus 50.1, but the 90% score intervals overlap. Treat that as a lead, not a settled winner.

4 results are shared. Category rows based on different benchmark sets are marked directional and do not name a winner.

Which one for your work

Recommendations appear only when a shared evidence basis or an explicit operating constraint supports the call. Secondary and unsupported use cases stay disclosed below the initial list.

Long documents
Prompts that approach the documented context limit
Qwen3.8 Max
Qwen3.8 Max has the larger documented context window.
Confidence: documented

Show secondary and unsupported calls

Coding work
Code generation, repair, and software-engineering tasks
Not enough matched evidence
The category averages use different weighted benchmark sets, so they are directional rather than like-for-like.
Confidence: limited
Agentic work
Tool use, computer use, and multi-step task completion
Not enough matched evidence
No shared weighted benchmark basis supports a winner.
Confidence: limited
Chat turn cost
1K fresh input + 500 output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-rates
Cache-heavy agent loop cost
200K cached + 20K fresh input + 10K output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-rates
Repository review cost
50K fresh input + 3K output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-rates

What is actually comparable

Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.

10 MAI-Thinking-1 only

4 Shared

48 Qwen3.8 Max only

Shared results: 4
MAI-Thinking-1 only: 10
Qwen3.8 Max only: 48
Like-for-like categories: 1 / 8

2 categories use different evidence sets. Those rows remain visible for coverage context but do not name a winner.

Category results, on a stated basis

Each row states whether both averages use the same weighted benchmark set. Directional and not-comparable rows remain visible, but they never receive a winner in this template.

Instruction following

Like-for-like

MAI-Thinking-1: 85.0
Qwen3.8 Max: 82.8
Weighted basis: 1 vs 1 rows
Reading: MAI-Thinking-1 leads

Coding

Directional only

MAI-Thinking-1: 65.5
Qwen3.8 Max: 67.7
Weighted basis: 2 vs 1 rows
Reading: Directional only

Knowledge

Directional only

MAI-Thinking-1: 72.5
Qwen3.8 Max: 50.2
Weighted basis: 3 vs 2 rows
Reading: Directional only

Agentic

Not comparable

MAI-Thinking-1: 46.0
Qwen3.8 Max: 86.1
Weighted basis: 1 vs 1 rows
Reading: Not comparable

Reasoning

Not comparable

MAI-Thinking-1: Not measured
Qwen3.8 Max: 78.3
Weighted basis: 0 vs 2 rows
Reading: Not comparable

Math

Not comparable

MAI-Thinking-1: 89.7
Qwen3.8 Max: Not measured
Weighted basis: 2 vs 0 rows
Reading: Not comparable

Multilingual

Not comparable

MAI-Thinking-1: Not measured
Qwen3.8 Max: Not measured
Weighted basis: 0 vs 0 rows
Reading: Not comparable

Multimodal

Not comparable

MAI-Thinking-1: Not measured
Qwen3.8 Max: 86.3
Weighted basis: 0 vs 2 rows
Reading: Not comparable

Category averages with the server-provided evidence basis for MAI-Thinking-1 and Qwen3.8 Max
Category	MAI-Thinking-1	Qwen3.8 Max	Weighted basis	Reading
Instruction following	85.0	82.8	Like-for-like1 vs 1 rows	MAI-Thinking-1 leads
Coding	65.5	67.7	Directional only2 vs 1 rows	Directional only
Knowledge	72.5	50.2	Directional only3 vs 2 rows	Directional only
Agentic	46.0	86.1	Not comparable1 vs 1 rows	Not comparable
Reasoning	Not measured	78.3	Not comparable0 vs 2 rows	Not comparable
Math	89.7	Not measured	Not comparable2 vs 0 rows	Not comparable
Multilingual	Not measured	Not measured	Not comparable0 vs 0 rows	Not comparable
Multimodal	Not measured	86.3	Not comparable0 vs 2 rows	Not comparable

Shape of the matched evidence

Only shared public evidence is shown. Sparse evidence stays a ruled list rather than being closed into a radar shape.

Too few matched category axes support a radar. The ruled list below shows only shared benchmark results; positions use each benchmark’s normalized display scale when available.

SWE-bench Pro
Coding
MAI-Thinking-1: 52.8%Qwen3.8 Max: 67.7%Normalized gap 14.9MAI-Thinking-1 source Qwen3.8 Max source
GPQA
Knowledge
MAI-Thinking-1: 84.2%Qwen3.8 Max: 92.6%Normalized gap 8.4MAI-Thinking-1 source Qwen3.8 Max source
IFBench
Instruction following
MAI-Thinking-1: 85%Qwen3.8 Max: 82.8%Normalized gap 2.2MAI-Thinking-1 source Qwen3.8 Max source

What each workload costs

Three fixed token mixes turn per-token rates into comparable decisions. Each scenario states context fit and whether cached input had to fall back to the published list-input rate.

Chat turn

1K fresh input + 500 output tokens

MAI-Thinking-1: API rate not published; Fits in one request
Qwen3.8 Max: API rate not published; Fits in one request

MAI-Thinking-1 has no comparable published API token rate. Qwen3.8 Max has no comparable published API token rate.

Repository review

50K fresh input + 3K output tokens

MAI-Thinking-1: API rate not published; Fits in one request
Qwen3.8 Max: API rate not published; Fits in one request

MAI-Thinking-1 has no comparable published API token rate. Qwen3.8 Max has no comparable published API token rate.

Cache-heavy agent loop

200K cached + 20K fresh input + 10K output tokens

MAI-Thinking-1: API rate not published; Fits in one request; Cached-input rate unavailable
Qwen3.8 Max: API rate not published; Fits in one request; Cached-input rate unavailable

MAI-Thinking-1 has no comparable published API token rate. Qwen3.8 Max has no comparable published API token rate.

Specification differences

Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.

SpecificationMAI-Thinking-1Qwen3.8 Max

Context window

Maximum documented context; output-token limits may be lower.

MAI-Thinking-1

256K

Qwen3.8 Max

Alibaba Cloud Model Studio pricing

API model ID

MAI-Thinking-1

Not sourced

Qwen3.8 Max

qwen3.8-max

Alibaba Cloud Model Studio pricing

Cached-input rate

A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.

MAI-Thinking-1

No comparable hosted API rate

Qwen3.8 Max

No comparable hosted API rate

Alibaba Cloud Model Studio pricing

Documented inputs

MAI-Thinking-1

Not sourced

Qwen3.8 Max

Not sourced

Documented outputs

MAI-Thinking-1

Not sourced

Qwen3.8 Max

Not sourced

Provider availability

MAI-Thinking-1

Not sourced

Qwen3.8 Max

Not sourced

Reasoning profile

MAI-Thinking-1

Reasoning

Qwen3.8 Max

Reasoning

Weight access

MAI-Thinking-1

Proprietary

Qwen3.8 Max

Proprietary

License

MAI-Thinking-1

Proprietary

Qwen3.8 Max

Proprietary

Release date

MAI-Thinking-1

2026-06-02

Qwen3.8 Max

2026-08-03

If you already use one of these models

Deployment change: The models list different providers, so authentication, endpoint behavior, limits, and feature support may change.
Quality signal: Qwen3.8 Max has the higher public score estimate, 65.4 versus 50.1, but the 90% score intervals overlap.
Workload cost: A complete comparable API-rate estimate is not available for both models.
Context tradeoff: Qwen3.8 Max has the larger documented window (1M).

Run the same representative tasks against both endpoints before changing production traffic.

Benchmark evidence

The full public result ledger is available for audit without forcing a wide desktop table onto a phone.

Browse raw public benchmark evidence62 rows

Agentic

Terminal-Bench 2.0
MAI-Thinking-146%
Source
Qwen3.8 Max—
Not directly comparable
Terminal-Bench 2.1
MAI-Thinking-1—
Qwen3.8 Max86.6%
Source
Not directly comparable
CoWorkBench
MAI-Thinking-1—
Qwen3.8 Max74.8%
Source
Not directly comparable
JobBench
MAI-Thinking-1—
Qwen3.8 Max53.4%
Source
Not directly comparable
skillsBench
MAI-Thinking-1—
Qwen3.8 Max70.2%
Source
Not directly comparable
Agents' Last Exam
MAI-Thinking-1—
Qwen3.8 Max52.4%
Source
Not directly comparable
AutomationBench
MAI-Thinking-1—
Qwen3.8 Max27.3%
Source
Not directly comparable
Toolathlon-Verified
MAI-Thinking-1—
Qwen3.8 Max72.5%
Source
Not directly comparable
WideResearch
MAI-Thinking-1—
Qwen3.8 Max81.9%
Source
Not directly comparable
HLE w/ tools
MAI-Thinking-1—
Qwen3.8 Max56.2%
Source
Not directly comparable
OSWorld-Verified
MAI-Thinking-1—
Qwen3.8 Max86.1%
Source
Not directly comparable
OSWorld 2.0
MAI-Thinking-1—
Qwen3.8 Max19.4%
Source
Not directly comparable
WebArena-Verified
MAI-Thinking-1—
Qwen3.8 Max66.8%
Source
Not directly comparable
AndroidWorld
MAI-Thinking-1—
Qwen3.8 Max85.3%
Source
Not directly comparable
MobileWorld
MAI-Thinking-1—
Qwen3.8 Max77.8%
Source
Not directly comparable

Coding

LiveCodeBench v6
MAI-Thinking-187.7%
Source
Qwen3.8 Max—
Not directly comparable
SWE-bench Verified
MAI-Thinking-173.5%
Source
Qwen3.8 Max—
Not directly comparable
SWE-bench Pro
MAI-Thinking-152.8%
Source
Qwen3.8 Max67.7%
Source
Qwen3.8 Max leads this result
Terminal-Bench 2.0
MAI-Thinking-146.0%
Source
Qwen3.8 Max—
Not directly comparable
Terminal-Bench 2.1
MAI-Thinking-1—
Qwen3.8 Max86.6%
Source
Not directly comparable
deepSwe
MAI-Thinking-1—
Qwen3.8 Max56.6%
Source
Not directly comparable
NL2Repo
MAI-Thinking-1—
Qwen3.8 Max55.9%
Source
Not directly comparable
FrontierSWE
MAI-Thinking-1—
Qwen3.8 Max73.5%
Source
Not directly comparable
MLS-Bench Lite
MAI-Thinking-1—
Qwen3.8 Max41.0%
Source
Not directly comparable
PaperBench
MAI-Thinking-1—
Qwen3.8 Max93.0%
Source
Not directly comparable

Reasoning

Graphwalks BFS 128K
MAI-Thinking-190%
Source
Qwen3.8 Max—
Not directly comparable
MRCRv2
MAI-Thinking-1—
Qwen3.8 Max92.9%
Source
Not directly comparable
LongBench v2
MAI-Thinking-1—
Qwen3.8 Max66.3%
Source
Not directly comparable

Knowledge

GPQA
MAI-Thinking-184.2%
Source
Qwen3.8 Max92.6%
Source
Qwen3.8 Max leads this result
GPQA-D
MAI-Thinking-184.2%
Source
Qwen3.8 Max92.6%
Source
Qwen3.8 Max leads this result
MMLU-Pro
MAI-Thinking-185%
Source
Qwen3.8 Max—
Not directly comparable
SimpleQA
MAI-Thinking-131%
Source
Qwen3.8 Max—
Not directly comparable
HLE
MAI-Thinking-1—
Qwen3.8 Max43.6%
Source
Not directly comparable
HLE w/o tools
MAI-Thinking-1—
Qwen3.8 Max43.6%
Source
Not directly comparable

Math

AIME 2025
MAI-Thinking-197%
Source
Qwen3.8 Max—
Not directly comparable
AIME26
MAI-Thinking-194.5%
Source
Qwen3.8 Max—
Not directly comparable
HMMT Feb 2026
MAI-Thinking-184.9%
Source
Qwen3.8 Max—
Not directly comparable

Multimodal

MMMU-Pro
MAI-Thinking-1—
Qwen3.8 Max82.3%
Source
Not directly comparable
MathVision
MAI-Thinking-1—
Qwen3.8 Max95.2%
Source
Not directly comparable
MathVision w/ Python
MAI-Thinking-1—
Qwen3.8 Max97.7%
Source
Not directly comparable
BabyVision
MAI-Thinking-1—
Qwen3.8 Max82.0%
Source
Not directly comparable
BabyVision w/ Python
MAI-Thinking-1—
Qwen3.8 Max91.3%
Source
Not directly comparable
ZeroBench
MAI-Thinking-1—
Qwen3.8 Max24.0%
Source
Not directly comparable
ZeroBench w/ Python
MAI-Thinking-1—
Qwen3.8 Max49.0%
Source
Not directly comparable
MedXpertQA (MM)
MAI-Thinking-1—
Qwen3.8 Max80.4%
Source
Not directly comparable
ScreenSpot Pro
MAI-Thinking-1—
Qwen3.8 Max84.5%
Source
Not directly comparable
Vision2Web
MAI-Thinking-1—
Qwen3.8 Max69.0%
Source
Not directly comparable
CharXiv w/o tools
MAI-Thinking-1—
Qwen3.8 Max88.4%
Source
Not directly comparable
CharXiv
MAI-Thinking-1—
Qwen3.8 Max93.5%
Source
Not directly comparable
OmniDocBench 1.5
MAI-Thinking-1—
Qwen3.8 Max92.1%
Source
Not directly comparable
OCRBench V2
MAI-Thinking-1—
Qwen3.8 Max74.2%
Source
Not directly comparable
CC-OCR
MAI-Thinking-1—
Qwen3.8 Max79.6%
Source
Not directly comparable
RealWorldQA
MAI-Thinking-1—
Qwen3.8 Max88.0%
Source
Not directly comparable
ERQA
MAI-Thinking-1—
Qwen3.8 Max77.8%
Source
Not directly comparable
SimpleVQA
MAI-Thinking-1—
Qwen3.8 Max75.0%
Source
Not directly comparable
PerceptionBench
MAI-Thinking-1—
Qwen3.8 Max63.5%
Source
Not directly comparable
Video-MME (with subtitle)
MAI-Thinking-1—
Qwen3.8 Max90.4%
Source
Not directly comparable
VideoMMMU
MAI-Thinking-1—
Qwen3.8 Max88.7%
Source
Not directly comparable
MMVU
MAI-Thinking-1—
Qwen3.8 Max82.4%
Source
Not directly comparable
MLVU (M-Avg)
MAI-Thinking-1—
Qwen3.8 Max90.8%
Source
Not directly comparable
LVBench
MAI-Thinking-1—
Qwen3.8 Max81.8%
Source
Not directly comparable

Instruction following

IFBench
MAI-Thinking-185%
Source
Qwen3.8 Max82.8%
Source
MAI-Thinking-1 leads this result

Frequently asked questions

Which is better, MAI-Thinking-1 or Qwen3.8 Max?

Qwen3.8 Max has the higher public score estimate, 65.4 versus 50.1, but the 90% score intervals overlap. The higher estimate is not a decisive winner because the uncertainty ranges overlap.

Which is better for coding, MAI-Thinking-1 or Qwen3.8 Max?

The current coding averages use different weighted benchmark sets, so BenchLM does not name a winner from them. Read the shared benchmark rows directly and test the models on the same task set.

Which is better for agentic tasks, MAI-Thinking-1 or Qwen3.8 Max?

The published evidence does not provide a shared weighted agentic tasks basis for both models, so BenchLM does not name a agentic tasks winner.

Which costs less, MAI-Thinking-1 or Qwen3.8 Max?

Both models do not have comparable published API token rates, so this page does not name a universal price winner.

Which has the larger context window, MAI-Thinking-1 or Qwen3.8 Max?

Qwen3.8 Max has the larger documented context window: 1M, compared with 256K.

Related comparisons

Compare API pricing Read the methodology Open the model selector

Last updated August 3, 2026

Watch MAI-Thinking-1 vs Qwen3.8 Max

One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.

Read a sample issue

Join 2,000+ readers.

MAI-Thinking-1 vs Qwen3.8 Max

Which one for your work

Long documents

Coding work

Agentic work

Chat turn cost

Cache-heavy agent loop cost

Repository review cost

What is actually comparable

Category results, on a stated basis

Instruction following

Coding

Knowledge

Agentic

Reasoning

Math

Multilingual

Multimodal

Shape of the matched evidence

What each workload costs

Chat turn

Repository review

Cache-heavy agent loop

Specification differences

Context window

API model ID

Cached-input rate

Documented inputs

Documented outputs

Provider availability

Reasoning profile

Weight access

License

Release date

Benchmark evidence

Agentic

Coding

Reasoning

Knowledge

Math

Multimodal

Instruction following

Frequently asked questions

Which is better, MAI-Thinking-1 or Qwen3.8 Max?

Which is better for coding, MAI-Thinking-1 or Qwen3.8 Max?

Which is better for agentic tasks, MAI-Thinking-1 or Qwen3.8 Max?

Which costs less, MAI-Thinking-1 or Qwen3.8 Max?

Which has the larger context window, MAI-Thinking-1 or Qwen3.8 Max?

Related comparisons

Watch MAI-Thinking-1 vs Qwen3.8 Max