Qwen3 Max Preview

Qwen3-Max-Preview shows substantial gains over the 2.5 series in overall capability, with significant enhancements in Chinese-English text understanding, complex instruction following, handling of su…

vercelAIGatewayCatalog/alibaba/qwen3-max-preview

Qwen3-Max-Preview shows substantial gains over the 2.5 series in overall capability, with significant enhancements in Chinese-English text understanding, complex instruction following, handling of subjective open-ended tasks, multilingual ability, and tool invocation; model knowledge hallucinations are reduced.

Specifications & capabilities

Input
text
Output
text
Context window · tokens
262144
Max output · tokens
32768

Evaluations

61–74 / 74

LMArena

1,399.288

source id
lmarena
leaderboard name
LMArena Text Factuality Creative Writing
dimension
factuality
metric
Arena rating
rank
112
sample size
496
evaluated at
2026-11-09T00:00:00.000Z
source model id
Qwen3-Max-Preview
confidence lower
1374.820571552434
confidence upper
1423.7561490272535
Original result ↗

LMArena

1,470.187

source id
lmarena
leaderboard name
LMArena Text Style Control Industry Software And It Services
dimension
style
metric
Arena rating
rank
113
sample size
9316
evaluated at
2026-11-09T00:00:00.000Z
source model id
Qwen3-Max-Preview
confidence lower
1463.5224385266424
confidence upper
1476.8507894592053
Original result ↗

LMArena

1,459.612

source id
lmarena
leaderboard name
LMArena Text Style Control Hard Prompts English
dimension
style
metric
Arena rating
rank
113
sample size
5809
evaluated at
2026-11-09T00:00:00.000Z
source model id
Qwen3-Max-Preview
confidence lower
1451.6458471897374
confidence upper
1467.5780071068207
Original result ↗

LMArena

1,424.914

source id
lmarena
leaderboard name
LMArena Text Style Control Russian
dimension
style
metric
Arena rating
rank
115
sample size
1864
evaluated at
2026-11-09T00:00:00.000Z
source model id
Qwen3-Max-Preview
confidence lower
1411.4129493135022
confidence upper
1438.4159811238494
Original result ↗

LMArena

1,480.426

source id
lmarena
leaderboard name
LMArena Text Style Control Coding
dimension
style
metric
Arena rating
rank
116
sample size
5356
evaluated at
2026-11-09T00:00:00.000Z
source model id
Qwen3-Max-Preview
confidence lower
1472.1078977923744
confidence upper
1488.7448437881485
Original result ↗

LMArena

1,440.274

source id
lmarena
leaderboard name
LMArena Text Style Control English
dimension
style
metric
Arena rating
rank
116
sample size
11454
evaluated at
2026-11-09T00:00:00.000Z
source model id
Qwen3-Max-Preview
confidence lower
1434.1577270682405
confidence upper
1446.3895786938674
Original result ↗

LMArena

1,397.418

source id
lmarena
leaderboard name
LMArena Text Style Control Industry Entertainment And Sports And Media
dimension
style
metric
Arena rating
rank
116
sample size
4943
evaluated at
2026-11-09T00:00:00.000Z
source model id
Qwen3-Max-Preview
confidence lower
1388.7376923781983
confidence upper
1406.097595988088
Original result ↗

LMArena

1,434.208

source id
lmarena
leaderboard name
LMArena Text Style Control Industry Mathematical
dimension
style
metric
Arena rating
rank
118
sample size
1323
evaluated at
2026-11-09T00:00:00.000Z
source model id
Qwen3-Max-Preview
confidence lower
1418.5315517299318
confidence upper
1449.8853036133946
Original result ↗

LMArena

1,405.565

source id
lmarena
leaderboard name
LMArena Text Style Control Industry Writing And Literature And Language
dimension
style
metric
Arena rating
rank
119
sample size
6164
evaluated at
2026-11-09T00:00:00.000Z
source model id
Qwen3-Max-Preview
confidence lower
1397.7248485092584
confidence upper
1413.4050018208652
Original result ↗

LMArena

1,393.967

source id
lmarena
leaderboard name
LMArena Text Factuality Industry Entertainment And Sports And Media
dimension
factuality
metric
Arena rating
rank
119
sample size
679
evaluated at
2026-11-09T00:00:00.000Z
source model id
Qwen3-Max-Preview
confidence lower
1373.0973126828865
confidence upper
1414.8371670769498
Original result ↗

LMArena

1,393.825

source id
lmarena
leaderboard name
LMArena Text Style Control Creative Writing
dimension
style
metric
Arena rating
rank
120
sample size
3715
evaluated at
2026-11-09T00:00:00.000Z
source model id
Qwen3-Max-Preview
confidence lower
1383.7918162678181
confidence upper
1403.8584921571219
Original result ↗

LMArena

1,405.313

source id
lmarena
leaderboard name
LMArena Text Factuality Industry Writing And Literature And Language
dimension
factuality
metric
Arena rating
rank
121
sample size
879
evaluated at
2026-11-09T00:00:00.000Z
source model id
Qwen3-Max-Preview
confidence lower
1387.553052896616
confidence upper
1423.0731584023747
Original result ↗

LMArena

1,413.288

source id
lmarena
leaderboard name
LMArena Text Factuality Instruction Following
dimension
factuality
metric
Arena rating
rank
127
sample size
1162
evaluated at
2026-11-09T00:00:00.000Z
source model id
Qwen3-Max-Preview
confidence lower
1397.4929256998287
confidence upper
1429.083573073652
Original result ↗

LMArena

1,431.496

source id
lmarena
leaderboard name
LMArena Text Style Control Industry Legal And Government
dimension
style
metric
Arena rating
rank
130
sample size
1689
evaluated at
2026-11-09T00:00:00.000Z
source model id
Qwen3-Max-Preview
confidence lower
1417.2480713622954
confidence upper
1445.7445421185332
Original result ↗

Safety assessments

1–1 / 1

Foundation Model Transparency Index 2025 ↗

26 / 100

source id
stanfordFMTI
canonical model id
vercelAIGatewayCatalog/alibaba/qwen3-max-preview
taxonomy
transparency
hazard category
transparency
hazard name
Organization transparency
policy area
Alibaba
assessment type
organization-level disclosure
metric
disclosure_score
result label
26 / 100
sample size
100
updated at
2026-11-09T18:08:05.227Z

Original sources

1–2 / 2

GitHub · Catalog ↗

LMArena · Catalog ↗

Source links

1–2 / 2

Explore more models

Models

Loading
EVIDENCE / COMPARISON

Benchmarks

Explore every benchmark

Loading

These charts contain model-linked results. The charts above also include unlinked source profiles. Compare scores within each evaluation. Model quality, usage, transparency and system performance measure different things.

Install TensorTrace