Agentic
Directional only- Qwen 3.6 Max (preview)
- 60.1
- Estimated · #26/151
- Qwen3.7 Plus
- 37.7
- Supported · #126/151
- Basis
- BenchAlign lane · 2 vs 11 public rows
- Reading
- Directional only
Qwen 3.6 Max (preview) vs Qwen3.7 Plus
Updated September 4, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.
Decision reading
Qwen 3.6 Max (preview) has the higher public score estimate, 63.64 versus 62.29, but the 90% score intervals overlap. Treat that as a lead, not a settled winner.
7 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
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
Qwen3.7 Plus
Qwen3.7 Plus 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
Qwen 3.6 Max (preview) and Qwen3.7 Plus are scored on Estimated evidence for coding, so the reading is directional rather than like-for-like.
Confidence: limited
Agentic work
Tool use, computer use, and multi-step task completion
Not enough matched evidence
Qwen 3.6 Max (preview) is scored on Estimated evidence for agentic, so the reading is directional rather than like-for-like.
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.
2 Qwen 3.6 Max (preview) only
7 Shared
45 Qwen3.7 Plus only
- Shared results
- 7
- Qwen 3.6 Max (preview) only
- 2
- Qwen3.7 Plus only
- 45
- Like-for-like categories
- 0 / 8
3 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.
Coding
Directional only- Qwen 3.6 Max (preview)
- 57.8
- Estimated · #33/183
- Qwen3.7 Plus
- 53.1
- Estimated · #51/183
- Basis
- BenchAlign lane · 4 vs 7 public rows
- Reading
- Directional only
Knowledge
Directional only- Qwen 3.6 Max (preview)
- 60.6
- Estimated · #40/181
- Qwen3.7 Plus
- 56.8
- Estimated · #50/181
- Basis
- BenchAlign lane · 1 vs 7 public rows
- Reading
- Directional only
Reasoning
Not comparable- Qwen 3.6 Max (preview)
- Not ranked
- Qwen3.7 Plus
- 73.7
- Unranked · 3 rankable rows
- Basis
- Provisional lane · 0 vs 1 weighted rows
- Reading
- Not comparable
Math
Not comparable- Qwen 3.6 Max (preview)
- 41.4
- Unranked · 2 rankable rows
- Qwen3.7 Plus
- 78.3
- Unranked · 3 rankable rows
- Basis
- Provisional lane · 2 vs 1 weighted rows
- Reading
- Not comparable
Multilingual
Not comparable- Qwen 3.6 Max (preview)
- Not ranked
- Qwen3.7 Plus
- 78.9
- #3/12
- Basis
- Provisional lane · 0 vs 1 weighted rows
- Reading
- Not comparable
Multimodal
Not comparable- Qwen 3.6 Max (preview)
- Not ranked
- Qwen3.7 Plus
- 72.5
- #18/48
- Basis
- Provisional lane · 0 vs 2 weighted rows
- Reading
- Not comparable
Instruction following
Not comparable- Qwen 3.6 Max (preview)
- Not ranked
- Qwen3.7 Plus
- 91.1
- #17/120
- Basis
- Provisional lane · 0 vs 1 weighted rows
- Reading
- Not comparable
| Category | Qwen 3.6 Max (preview) | Qwen3.7 Plus | Basis | Reading |
|---|---|---|---|---|
| Agentic | 60.1Estimated · #26/151 | 37.7Supported · #126/151 | Directional onlyBenchAlign lane · 2 vs 11 public rows | Directional only |
| Coding | 57.8Estimated · #33/183 | 53.1Estimated · #51/183 | Directional onlyBenchAlign lane · 4 vs 7 public rows | Directional only |
| Knowledge | 60.6Estimated · #40/181 | 56.8Estimated · #50/181 | Directional onlyBenchAlign lane · 1 vs 7 public rows | Directional only |
| Reasoning | Not ranked | 73.7Unranked · 3 rankable rows | Not comparableProvisional lane · 0 vs 1 weighted rows | Not comparable |
| Math | 41.4Unranked · 2 rankable rows | 78.3Unranked · 3 rankable rows | Not comparableProvisional lane · 2 vs 1 weighted rows | Not comparable |
| Multilingual | Not ranked | 78.9#3/12 | Not comparableProvisional lane · 0 vs 1 weighted rows | Not comparable |
| Multimodal | Not ranked | 72.5#18/48 | Not comparableProvisional lane · 0 vs 2 weighted rows | Not comparable |
| Instruction following | Not ranked | 91.1#17/120 | Not comparableProvisional lane · 0 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
Qwen 3.6 Max (preview): 65.4%Qwen3.7 Plus: 70.3%Normalized gap 4.9Qwen 3.6 Max (preview) source Qwen3.7 Plus sourceSciCode
Coding
Qwen 3.6 Max (preview): 47%Qwen3.7 Plus: 51.3%Normalized gap 4.3Qwen 3.6 Max (preview) source Qwen3.7 Plus sourceSuperGPQA
Knowledge
Qwen 3.6 Max (preview): 73.9%Qwen3.7 Plus: 71.4%Normalized gap 2.5Qwen 3.6 Max (preview) source Qwen3.7 Plus sourceSWE-bench Pro
Coding
Qwen 3.6 Max (preview): 57.3%Qwen3.7 Plus: 57.6%Normalized gap 0.3Qwen 3.6 Max (preview) 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
- Qwen 3.6 Max (preview)
- API rate not published
- Fits in one request
- Qwen3.7 Plus
- API rate not published
- Fits in one request
Qwen 3.6 Max (preview) 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
- Qwen 3.6 Max (preview)
- API rate not published
- Fits in one request
- Qwen3.7 Plus
- API rate not published
- Fits in one request
Qwen 3.6 Max (preview) 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
- Qwen 3.6 Max (preview)
- 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
Qwen 3.6 Max (preview) 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.
SpecificationQwen 3.6 Max (preview)Qwen3.7 Plus
Context window
Maximum documented context; output-token limits may be lower.
Qwen 3.6 Max (preview)
256K
Qwen3.7 Plus
1M
API model ID
Qwen 3.6 Max (preview)
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.
Qwen 3.6 Max (preview)
No comparable hosted API rate
Qwen3.7 Plus
No comparable hosted API rate
Documented inputs
Qwen 3.6 Max (preview)
Not sourced
Qwen3.7 Plus
Not sourced
Documented outputs
Qwen 3.6 Max (preview)
Not sourced
Qwen3.7 Plus
Not sourced
Provider availability
Qwen 3.6 Max (preview)
Not sourced
Qwen3.7 Plus
Not sourced
Reasoning profile
Qwen 3.6 Max (preview)
Reasoning
Qwen3.7 Plus
Reasoning
Weight access
Qwen 3.6 Max (preview)
Proprietary
Qwen3.7 Plus
Proprietary
License
Qwen 3.6 Max (preview)
Proprietary
Qwen3.7 Plus
Proprietary
Release date
Qwen 3.6 Max (preview)
2026-04-20
Qwen3.7 Plus
2026-06-03
If you already use one of these models
- Deployment change
- Both entries list Alibaba as the provider. Confirm endpoint, model ID, limits, and feature support before switching.
- Quality signal
- Qwen 3.6 Max (preview) has the higher public score estimate, 63.64 versus 62.29, but the 90% score intervals overlap.
- Workload cost
- A complete comparable API-rate estimate is not available for both models.
- Context tradeoff
- Qwen3.7 Plus 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 evidence54 rows
Agentic
Terminal-Bench 2.0
Qwen 3.6 Max (preview)65.4%SourceQwen3.7 Plus70.3%SourceQwen3.7 Plus leads this result
QwenClawBench
Qwen 3.6 Max (preview)59.0%SourceQwen3.7 Plus61.8%SourceQwen3.7 Plus leads this result
Claw-Eval
Qwen 3.6 Max (preview)—Qwen3.7 Plus62.7%SourceNot directly comparable
BFCL v4
Qwen 3.6 Max (preview)—Qwen3.7 Plus72.9%SourceNot directly comparable
MCP Atlas
Qwen 3.6 Max (preview)—Qwen3.7 Plus73.2%SourceNot directly comparable
VITA-Bench
Qwen 3.6 Max (preview)—Qwen3.7 Plus45.6%SourceNot directly comparable
DeepPlanning
Qwen 3.6 Max (preview)—Qwen3.7 Plus62.3%SourceNot directly comparable
OSWorld-Verified
Qwen 3.6 Max (preview)—Qwen3.7 Plus73.3%SourceNot directly comparable
AndroidWorld
Qwen 3.6 Max (preview)—Qwen3.7 Plus81.0%SourceNot directly comparable
OSWorld 2.0
Qwen 3.6 Max (preview)—Qwen3.7 Plus2.8%SourceNot directly comparable
Terminal-Bench 2.1 (Vals)
Qwen 3.6 Max (preview)—Qwen3.7 Plus52.8%SourceNot directly comparable
Coding
SWE-bench Pro
Qwen 3.6 Max (preview)57.3%SourceQwen3.7 Plus57.6%SourceQwen3.7 Plus leads this result
SciCode
Qwen 3.6 Max (preview)47%SourceQwen3.7 Plus51.3%SourceQwen3.7 Plus leads this result
NL2Repo
Qwen 3.6 Max (preview)42.9%SourceQwen3.7 Plus41.1%SourceQwen 3.6 Max (preview) leads this result
Terminal-Bench 2.0
Qwen 3.6 Max (preview)65.4%SourceQwen3.7 Plus70.3%SourceQwen3.7 Plus leads this result
SWE-bench Verified
Qwen 3.6 Max (preview)—Qwen3.7 Plus77.7%SourceNot directly comparable
SWE Multilingual
Qwen 3.6 Max (preview)—Qwen3.7 Plus75.8%SourceNot directly comparable
LiveCodeBench
Qwen 3.6 Max (preview)—Qwen3.7 Plus89.6%SourceNot directly comparable
Reasoning
Knowledge
SuperGPQA
Qwen 3.6 Max (preview)73.9%SourceQwen3.7 Plus71.4%SourceQwen 3.6 Max (preview) leads this result
GPQA
Qwen 3.6 Max (preview)—Qwen3.7 Plus90.3%SourceNot directly comparable
GPQA-D
Qwen 3.6 Max (preview)—Qwen3.7 Plus90.3%SourceNot directly comparable
HLE
Qwen 3.6 Max (preview)—Qwen3.7 Plus34.7%SourceNot directly comparable
MMLU-Pro
Qwen 3.6 Max (preview)—Qwen3.7 Plus88.5%SourceNot directly comparable
MMLU-Redux
Qwen 3.6 Max (preview)—Qwen3.7 Plus94.5%SourceNot directly comparable
MMMLU
Qwen 3.6 Max (preview)—Qwen3.7 Plus89.0%SourceNot directly comparable
Math
FrontierMath v2 (Tiers 1-3)
Qwen 3.6 Max (preview)23.103%SourceQwen3.7 Plus—Not directly comparable
FrontierMath v2 (Tier 4)
Qwen 3.6 Max (preview)4.167%SourceQwen3.7 Plus—Not directly comparable
HMMT Feb 2026
Qwen 3.6 Max (preview)—Qwen3.7 Plus92.9%SourceNot directly comparable
IMOAnswerBench
Qwen 3.6 Max (preview)—Qwen3.7 Plus86.0%SourceNot directly comparable
Apex
Qwen 3.6 Max (preview)—Qwen3.7 Plus22.7%SourceNot directly comparable
Multilingual
MMLU-ProX
Qwen 3.6 Max (preview)—Qwen3.7 Plus85.4%SourceNot directly comparable
NOVA-63
Qwen 3.6 Max (preview)—Qwen3.7 Plus58.8%SourceNot directly comparable
INCLUDE
Qwen 3.6 Max (preview)—Qwen3.7 Plus83.0%SourceNot directly comparable
MAXIFE
Qwen 3.6 Max (preview)—Qwen3.7 Plus88.8%SourceNot directly comparable
PolyMath
Qwen 3.6 Max (preview)—Qwen3.7 Plus84.0%SourceNot directly comparable
Multimodal
MMMU-Pro
Qwen 3.6 Max (preview)—Qwen3.7 Plus79%SourceNot directly comparable
MathVision
Qwen 3.6 Max (preview)—Qwen3.7 Plus90.3%SourceNot directly comparable
CharXiv
Qwen 3.6 Max (preview)—Qwen3.7 Plus85.9%SourceNot directly comparable
ERQA
Qwen 3.6 Max (preview)—Qwen3.7 Plus69.8%SourceNot directly comparable
MedXpertQA (MM)
Qwen 3.6 Max (preview)—Qwen3.7 Plus71.0%SourceNot directly comparable
ScreenSpot Pro
Qwen 3.6 Max (preview)—Qwen3.7 Plus79.0%SourceNot directly comparable
SimpleVQA
Qwen 3.6 Max (preview)—Qwen3.7 Plus81.7%SourceNot directly comparable
MMSearch-Plus
Qwen 3.6 Max (preview)—Qwen3.7 Plus41.4%SourceNot directly comparable
RealWorldQA
Qwen 3.6 Max (preview)—Qwen3.7 Plus86.9%SourceNot directly comparable
OmniDocBench 1.5
Qwen 3.6 Max (preview)—Qwen3.7 Plus91.4%SourceNot directly comparable
OCRBench V2
Qwen 3.6 Max (preview)—Qwen3.7 Plus70.7%SourceNot directly comparable
ODINW13
Qwen 3.6 Max (preview)—Qwen3.7 Plus51.1%SourceNot directly comparable
Video-MME (with subtitle)
Qwen 3.6 Max (preview)—Qwen3.7 Plus88.0%SourceNot directly comparable
VideoMMMU
Qwen 3.6 Max (preview)—Qwen3.7 Plus85.4%SourceNot directly comparable
MLVU (M-Avg)
Qwen 3.6 Max (preview)—Qwen3.7 Plus87.4%SourceNot directly comparable
Instruction following
Frequently asked questions
Which is better, Qwen 3.6 Max (preview) or Qwen3.7 Plus?
Qwen 3.6 Max (preview) has the higher public score estimate, 63.64 versus 62.29, but the 90% score intervals overlap. The higher estimate is not a decisive winner because the uncertainty ranges overlap.
Which is better for coding, Qwen 3.6 Max (preview) or Qwen3.7 Plus?
Qwen 3.6 Max (preview) scores higher for coding on the public lane, 57.8 to 53.1. Qwen 3.6 Max (preview) and Qwen3.7 Plus are 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, Qwen 3.6 Max (preview) or Qwen3.7 Plus?
Qwen 3.6 Max (preview) scores higher for agentic tasks on the public lane, 60.1 to 37.7. Qwen 3.6 Max (preview) is scored on Estimated evidence for agentic tasks, so the reading is directional rather than like-for-like. BenchLM does not name a winner for agentic tasks from a directional reading; read the shared benchmark rows directly and test the models on the same task set.
Which costs less, Qwen 3.6 Max (preview) 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, Qwen 3.6 Max (preview) or Qwen3.7 Plus?
Qwen3.7 Plus has the larger documented context window: 1M, compared with 256K.
Related comparisons
Last updated September 4, 2026
Watch Qwen 3.6 Max (preview) vs Qwen3.7 Plus
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 2,000+ readers.