Model A

Alibaba

61.49/100

Supported · Public rank #64

90% interval 53.0–69.9

Qwen3.6 Plus vs Qwen3.7 Plus

Updated September 4, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.

Model B

Qwen3.7 Plus

Alibaba

62.29/100

Supported · Public rank #60

90% interval 52.4–72.2

Decision reading

Qwen3.7 Plus has the higher public score estimate, 62.29 versus 61.49, but the 90% score intervals overlap. Treat that as a lead, not a settled winner.

25 results are shared. Category rows resting on Estimated evidence or different benchmark sets are marked directional and do not name a winner.

Share or export

Share on X LinkedInSocial card CSV JSON

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.

Agentic work
Tool use, computer use, and multi-step task completion
Qwen3.7 Plus
Qwen3.7 Plus leads on the public agentic lane, 37.7 to 36.4, with Supported evidence for both models, although the 90% intervals overlap.
Confidence: limited

Show secondary and unsupported calls

Coding work
Code generation, repair, and software-engineering tasks
Not enough matched evidence
Qwen3.7 Plus is scored on Estimated evidence for coding, so the reading is directional rather than like-for-like.
Confidence: limited
Long documents
Prompts that approach the documented context limit
No clear pick
The documented context windows are equal.
Confidence: documented
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.

23 Qwen3.6 Plus only

25 Shared

27 Qwen3.7 Plus only

Shared results: 25
Qwen3.6 Plus only: 23
Qwen3.7 Plus only: 27
Like-for-like categories: 4 / 8

2 categories rest on Estimated evidence or different evidence sets. Those rows remain visible for coverage context but do not name a winner.

Category results, on a stated basis

Each row shows the public-lane category score for both models: the BenchAlign lane for agentic, coding, and knowledge, and the weighted public rows elsewhere, with the same rank each model page reports. A row is like-for-like only when both scores rest on Supported evidence or the same weighted set. Directional and not-comparable rows remain visible, but they never receive a winner.

Agentic

Like-for-like

Qwen3.6 Plus: 36.4; Supported · #128/151
Qwen3.7 Plus: 37.7; Supported · #126/151
Basis: BenchAlign lane · 13 vs 11 public rows
Reading: Qwen3.7 Plus leads · intervals overlap

Multilingual

Like-for-like

Qwen3.6 Plus: 69.7; #4/12
Qwen3.7 Plus: 78.9; #3/12
Basis: Provisional lane · 1 vs 1 weighted rows
Reading: Qwen3.7 Plus leads

Multimodal

Like-for-like

Qwen3.6 Plus: 66.3; #22/48
Qwen3.7 Plus: 72.5; #18/48
Basis: Provisional lane · 2 vs 2 weighted rows
Reading: Qwen3.7 Plus leads

Instruction following

Like-for-like

Qwen3.6 Plus: 85.5; #41/120
Qwen3.7 Plus: 91.1; #17/120
Basis: Provisional lane · 1 vs 1 weighted rows
Reading: Qwen3.7 Plus leads

Coding

Directional only

Qwen3.6 Plus: 47.0; Supported · #92/183
Qwen3.7 Plus: 53.1; Estimated · #51/183
Basis: BenchAlign lane · 7 vs 7 public rows
Reading: Directional only

Knowledge

Directional only

Qwen3.6 Plus: 54.7; Supported · #63/181
Qwen3.7 Plus: 56.8; Estimated · #50/181
Basis: BenchAlign lane · 8 vs 7 public rows
Reading: Directional only

Reasoning

Not comparable

Qwen3.6 Plus: 59.3; Unranked · 4 rankable rows
Qwen3.7 Plus: 73.7; Unranked · 3 rankable rows
Basis: Provisional lane · 1 vs 1 weighted rows
Reading: Not comparable

Math

Not comparable

Qwen3.6 Plus: 62.4; #4/7
Qwen3.7 Plus: 78.3; Unranked · 3 rankable rows
Basis: Provisional lane · 4 vs 1 weighted rows
Reading: Not comparable

Category scores with the server-provided evidence basis for Qwen3.6 Plus and Qwen3.7 Plus
Category	Qwen3.6 Plus	Qwen3.7 Plus	Basis	Reading
Agentic	36.4Supported · #128/151	37.7Supported · #126/151	Like-for-likeBenchAlign lane · 13 vs 11 public rows	Qwen3.7 Plus leads · intervals overlap
Multilingual	69.7#4/12	78.9#3/12	Like-for-likeProvisional lane · 1 vs 1 weighted rows	Qwen3.7 Plus leads
Multimodal	66.3#22/48	72.5#18/48	Like-for-likeProvisional lane · 2 vs 2 weighted rows	Qwen3.7 Plus leads
Instruction following	85.5#41/120	91.1#17/120	Like-for-likeProvisional lane · 1 vs 1 weighted rows	Qwen3.7 Plus leads
Coding	47.0Supported · #92/183	53.1Estimated · #51/183	Directional onlyBenchAlign lane · 7 vs 7 public rows	Directional only
Knowledge	54.7Supported · #63/181	56.8Estimated · #50/181	Directional onlyBenchAlign lane · 8 vs 7 public rows	Directional only
Reasoning	59.3Unranked · 4 rankable rows	73.7Unranked · 3 rankable rows	Not comparableProvisional lane · 1 vs 1 weighted rows	Not comparable
Math	62.4#4/7	78.3Unranked · 3 rankable rows	Not comparableProvisional lane · 4 vs 1 weighted rows	Not comparable

Ranks count the models scored in each category’s lane, so the agentic, coding, and knowledge denominators (BenchAlign) differ from the provisional-lane categories. Unranked scores sit on the lane’s scale but never name a winner. Like-for-like rows within 0.5 points read as a practical tie.

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.

Terminal-Bench 2.0
Agentic
Qwen3.6 Plus: 61.6%Qwen3.7 Plus: 70.3%Normalized gap 8.7Qwen3.6 Plus source Qwen3.7 Plus source
HLE
Knowledge
Qwen3.6 Plus: 28.8%Qwen3.7 Plus: 34.7%Normalized gap 5.9Qwen3.6 Plus source Qwen3.7 Plus source
HMMT Feb 2026
Math
Qwen3.6 Plus: 87.8%Qwen3.7 Plus: 92.9%Normalized gap 5.1Qwen3.6 Plus source Qwen3.7 Plus source
CharXiv
Multimodal
Qwen3.6 Plus: 81.5%Qwen3.7 Plus: 85.9%Normalized gap 4.4Qwen3.6 Plus source Qwen3.7 Plus source
IFBench
Instruction following
Qwen3.6 Plus: 75.8%Qwen3.7 Plus: 79.1%Normalized gap 3.3Qwen3.6 Plus source Qwen3.7 Plus 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

Qwen3.6 Plus: API rate not published; Fits in one request
Qwen3.7 Plus: API rate not published; Fits in one request

Qwen3.6 Plus has no comparable published API token rate. Qwen3.7 Plus has no comparable published API token rate.

Repository review

50K fresh input + 3K output tokens

Qwen3.6 Plus: API rate not published; Fits in one request
Qwen3.7 Plus: API rate not published; Fits in one request

Qwen3.6 Plus has no comparable published API token rate. Qwen3.7 Plus has no comparable published API token rate.

Cache-heavy agent loop

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

Qwen3.6 Plus: API rate not published; Fits in one request; Cached-input rate unavailable
Qwen3.7 Plus: API rate not published; Fits in one request; Cached-input rate unavailable

Qwen3.6 Plus has no comparable published API token rate. Qwen3.7 Plus 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.

SpecificationQwen3.6 PlusQwen3.7 Plus

Context window

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

Qwen3.6 Plus

Qwen3.7 Plus

API model ID

Qwen3.6 Plus

Not sourced

Qwen3.7 Plus

Not sourced

Cached-input rate

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

Qwen3.6 Plus

No comparable hosted API rate

Qwen3.7 Plus

No comparable hosted API rate

Documented inputs

Qwen3.6 Plus

Not sourced

Qwen3.7 Plus

Not sourced

Documented outputs

Qwen3.6 Plus

Not sourced

Qwen3.7 Plus

Not sourced

Provider availability

Qwen3.6 Plus

Not sourced

Qwen3.7 Plus

Not sourced

Reasoning profile

Qwen3.6 Plus

Reasoning

Qwen3.7 Plus

Reasoning

Weight access

Qwen3.6 Plus

Proprietary

Qwen3.7 Plus

Proprietary

License

Qwen3.6 Plus

Proprietary

Qwen3.7 Plus

Proprietary

Release date

Qwen3.6 Plus

2026-04-02

Qwen3.7 Plus

2026-06-03

If you are considering the documented upgrade path

Deployment change: Both entries list Alibaba as the provider. Confirm endpoint, model ID, limits, and feature support before switching.
Quality signal: Qwen3.7 Plus has the higher public score estimate, 62.29 versus 61.49, but the 90% score intervals overlap.
Workload cost: A complete comparable API-rate estimate is not available for both models.
Context tradeoff: Both models list 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 evidence75 rows

Agentic

Terminal-Bench 2.0
Qwen3.6 Plus61.6%
Source
Qwen3.7 Plus70.3%
Source
Qwen3.7 Plus leads this result
Claw-Eval
Qwen3.6 Plus58.8%
Source
Qwen3.7 Plus62.7%
Source
Qwen3.7 Plus leads this result
QwenClawBench
Qwen3.6 Plus57.2%
Source
Qwen3.7 Plus61.8%
Source
Qwen3.7 Plus leads this result
τ³-bench results
Qwen3.6 Plus70.7%
Source
Qwen3.7 Plus—
Not directly comparable
VITA-Bench
Qwen3.6 Plus44.3%
Source
Qwen3.7 Plus45.6%
Source
Qwen3.7 Plus leads this result
DeepPlanning
Qwen3.6 Plus41.5%
Source
Qwen3.7 Plus62.3%
Source
Qwen3.7 Plus leads this result
Toolathlon
Qwen3.6 Plus39.8%
Source
Qwen3.7 Plus—
Not directly comparable
MCP Atlas
Qwen3.6 Plus48.2%
Source
Qwen3.7 Plus73.2%
Source
Qwen3.7 Plus leads this result
MCP-Tasks
Qwen3.6 Plus74.1%
Source
Qwen3.7 Plus—
Not directly comparable
WideResearch
Qwen3.6 Plus74.3%
Source
Qwen3.7 Plus—
Not directly comparable
Gert Labs
Qwen3.6 Plus50.60%
Source
Qwen3.7 Plus—
Not directly comparable
ResearchClawBench
Qwen3.6 Plus18.0%
Source
Qwen3.7 Plus—
Not directly comparable
Terminal-Bench 2.1 (Vals)
Qwen3.6 Plus53.2%
Source
Qwen3.7 Plus52.8%
Source
Qwen3.6 Plus leads this result
BFCL v4
Qwen3.6 Plus—
Qwen3.7 Plus72.9%
Source
Not directly comparable
OSWorld-Verified
Qwen3.6 Plus—
Qwen3.7 Plus73.3%
Source
Not directly comparable
AndroidWorld
Qwen3.6 Plus—
Qwen3.7 Plus81.0%
Source
Not directly comparable
OSWorld 2.0
Qwen3.6 Plus—
Qwen3.7 Plus2.8%
Source
Not directly comparable

Coding

SWE-bench Verified
Qwen3.6 Plus78.8%
Source
Qwen3.7 Plus77.7%
Source
Qwen3.6 Plus leads this result
SWE-bench Pro
Qwen3.6 Plus56.6%
Source
Qwen3.7 Plus57.6%
Source
Qwen3.7 Plus leads this result
SWE Multilingual
Qwen3.6 Plus73.8%
Source
Qwen3.7 Plus75.8%
Source
Qwen3.7 Plus leads this result
LiveCodeBench v6
Qwen3.6 Plus87.1%
Source
Qwen3.7 Plus—
Not directly comparable
Vibe Code Bench
Qwen3.6 Plus25.56%
Source
Qwen3.7 Plus—
Not directly comparable
LiveCodeBench (Vals)
Qwen3.6 Plus86.0%
Source
Qwen3.7 Plus—
Not directly comparable
SWE-bench (Vals)
Qwen3.6 Plus73.4%
Source
Qwen3.7 Plus—
Not directly comparable
Terminal-Bench 2.0
Qwen3.6 Plus—
Qwen3.7 Plus70.3%
Source
Not directly comparable
NL2Repo
Qwen3.6 Plus—
Qwen3.7 Plus41.1%
Source
Not directly comparable
SciCode
Qwen3.6 Plus—
Qwen3.7 Plus51.3%
Source
Not directly comparable
LiveCodeBench
Qwen3.6 Plus—
Qwen3.7 Plus89.6%
Source
Not directly comparable

Reasoning

AI-Needle
Qwen3.6 Plus68.3%
Source
Qwen3.7 Plus—
Not directly comparable
LongBench v2
Qwen3.6 Plus62%
Source
Qwen3.7 Plus—
Not directly comparable
CritPt
Qwen3.6 Plus—
Qwen3.7 Plus9.1%
Source
Not directly comparable
MRCRv2
Qwen3.6 Plus—
Qwen3.7 Plus91.7%
Source
Not directly comparable

Knowledge

GPQA
Qwen3.6 Plus90.4%
Source
Qwen3.7 Plus90.3%
Source
Qwen3.6 Plus leads this result
SuperGPQA
Qwen3.6 Plus71.6%
Source
Qwen3.7 Plus71.4%
Source
Qwen3.6 Plus leads this result
MMLU-Pro
Qwen3.6 Plus88.5%
Source
Qwen3.7 Plus88.5%
Source
Tie
MMLU-Redux
Qwen3.6 Plus94.5%
Source
Qwen3.7 Plus94.5%
Source
Tie
C-Eval
Qwen3.6 Plus93.3%
Source
Qwen3.7 Plus—
Not directly comparable
HLE
Qwen3.6 Plus28.8%
Source
Qwen3.7 Plus34.7%
Source
Qwen3.7 Plus leads this result
GPQA Diamond (Vals)
Qwen3.6 Plus87.4%
Source
Qwen3.7 Plus—
Not directly comparable
MMLU-Pro (Vals)
Qwen3.6 Plus87.7%
Source
Qwen3.7 Plus—
Not directly comparable
GPQA-D
Qwen3.6 Plus—
Qwen3.7 Plus90.3%
Source
Not directly comparable
MMMLU
Qwen3.6 Plus—
Qwen3.7 Plus89.0%
Source
Not directly comparable

Math

AIME26
Qwen3.6 Plus95.3%
Source
Qwen3.7 Plus—
Not directly comparable
HMMT Feb 2025
Qwen3.6 Plus96.7%
Source
Qwen3.7 Plus—
Not directly comparable
HMMT Nov 2025
Qwen3.6 Plus94.6%
Source
Qwen3.7 Plus—
Not directly comparable
HMMT Feb 2026
Qwen3.6 Plus87.8%
Source
Qwen3.7 Plus92.9%
Source
Qwen3.7 Plus leads this result
MMAnswerBench
Qwen3.6 Plus83.8%
Source
Qwen3.7 Plus—
Not directly comparable
FrontierMath v2 (Tiers 1-3)
Qwen3.6 Plus26.207%
Source
Qwen3.7 Plus—
Not directly comparable
FrontierMath v2 (Tier 4)
Qwen3.6 Plus8.333%
Source
Qwen3.7 Plus—
Not directly comparable
IMOAnswerBench
Qwen3.6 Plus—
Qwen3.7 Plus86.0%
Source
Not directly comparable
Apex
Qwen3.6 Plus—
Qwen3.7 Plus22.7%
Source
Not directly comparable

Multilingual

MMLU-ProX
Qwen3.6 Plus84.7%
Source
Qwen3.7 Plus85.4%
Source
Qwen3.7 Plus leads this result
NOVA-63
Qwen3.6 Plus57.9%
Source
Qwen3.7 Plus58.8%
Source
Qwen3.7 Plus leads this result
INCLUDE
Qwen3.6 Plus—
Qwen3.7 Plus83.0%
Source
Not directly comparable
MAXIFE
Qwen3.6 Plus—
Qwen3.7 Plus88.8%
Source
Not directly comparable
PolyMath
Qwen3.6 Plus—
Qwen3.7 Plus84.0%
Source
Not directly comparable

Multimodal

MMMU
Qwen3.6 Plus86.0%
Source
Qwen3.7 Plus—
Not directly comparable
MMMU-Pro
Qwen3.6 Plus78.8%
Source
Qwen3.7 Plus79%
Source
Qwen3.7 Plus leads this result
MathVision
Qwen3.6 Plus88.0%
Source
Qwen3.7 Plus90.3%
Source
Qwen3.7 Plus leads this result
VideoMMMU
Qwen3.6 Plus84.0%
Source
Qwen3.7 Plus85.4%
Source
Qwen3.7 Plus leads this result
ScreenSpot Pro
Qwen3.6 Plus68.2%
Source
Qwen3.7 Plus79.0%
Source
Qwen3.7 Plus leads this result
CharXiv
Qwen3.6 Plus81.5%
Source
Qwen3.7 Plus85.9%
Source
Qwen3.7 Plus leads this result
V*
Qwen3.6 Plus96.9%
Source
Qwen3.7 Plus—
Not directly comparable
ERQA
Qwen3.6 Plus—
Qwen3.7 Plus69.8%
Source
Not directly comparable
MedXpertQA (MM)
Qwen3.6 Plus—
Qwen3.7 Plus71.0%
Source
Not directly comparable
SimpleVQA
Qwen3.6 Plus—
Qwen3.7 Plus81.7%
Source
Not directly comparable
MMSearch-Plus
Qwen3.6 Plus—
Qwen3.7 Plus41.4%
Source
Not directly comparable
RealWorldQA
Qwen3.6 Plus—
Qwen3.7 Plus86.9%
Source
Not directly comparable
OmniDocBench 1.5
Qwen3.6 Plus—
Qwen3.7 Plus91.4%
Source
Not directly comparable
OCRBench V2
Qwen3.6 Plus—
Qwen3.7 Plus70.7%
Source
Not directly comparable
ODINW13
Qwen3.6 Plus—
Qwen3.7 Plus51.1%
Source
Not directly comparable
Video-MME (with subtitle)
Qwen3.6 Plus—
Qwen3.7 Plus88.0%
Source
Not directly comparable
MLVU (M-Avg)
Qwen3.6 Plus—
Qwen3.7 Plus87.4%
Source
Not directly comparable

Instruction following

IFEval
Qwen3.6 Plus94.3%
Source
Qwen3.7 Plus94.6%
Source
Qwen3.7 Plus leads this result
IFBench
Qwen3.6 Plus75.8%
Source
Qwen3.7 Plus79.1%
Source
Qwen3.7 Plus leads this result

Frequently asked questions

Which is better, Qwen3.6 Plus or Qwen3.7 Plus?

Qwen3.7 Plus has the higher public score estimate, 62.29 versus 61.49, but the 90% score intervals overlap. The higher estimate is not a decisive winner because the uncertainty ranges overlap.

Which is better for coding, Qwen3.6 Plus or Qwen3.7 Plus?

Qwen3.7 Plus scores higher for coding on the public lane, 53.1 to 47. Qwen3.7 Plus is scored on Estimated evidence for coding, so the reading is directional rather than like-for-like. BenchLM does not name a winner for coding from a directional reading; read the shared benchmark rows directly and test the models on the same task set.

Which is better for agentic tasks, Qwen3.6 Plus or Qwen3.7 Plus?

Qwen3.7 Plus leads the public agentic tasks lane, 37.7 to 36.4, with Supported evidence for both models, although the 90% intervals overlap.

Which costs less, Qwen3.6 Plus or Qwen3.7 Plus?

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, Qwen3.6 Plus or Qwen3.7 Plus?

Both models list the same context window, 1M.

Related comparisons

Compare API pricing Read the methodology Open the model selector

Last updated September 4, 2026

Watch Qwen3.6 Plus vs Qwen3.7 Plus

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

Read a sample issue

Join 2,000+ readers.

Qwen3.6 Plus vs Qwen3.7 Plus

Which one for your work

Agentic work

Coding work

Long documents

Chat turn cost

Cache-heavy agent loop cost

Repository review cost

What is actually comparable

Category results, on a stated basis

Agentic

Multilingual

Multimodal

Instruction following

Coding

Knowledge

Reasoning

Math

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

Multilingual

Multimodal

Instruction following

Frequently asked questions

Which is better, Qwen3.6 Plus or Qwen3.7 Plus?

Which is better for coding, Qwen3.6 Plus or Qwen3.7 Plus?

Which is better for agentic tasks, Qwen3.6 Plus or Qwen3.7 Plus?

Which costs less, Qwen3.6 Plus or Qwen3.7 Plus?

Which has the larger context window, Qwen3.6 Plus or Qwen3.7 Plus?

Related comparisons

Watch Qwen3.6 Plus vs Qwen3.7 Plus