Qwen3.5 397B (Reasoning) vs Qwen3 Max
Change modelsQwen3.5 397B (Reasoning)Alibaba
Qwen3 MaxAlibaba
IndexAll categories, out of 100.
55.3 ±6.9 best
45.0 ±5.1
GapIndex points below the top.
Top
−10.3
CoverageIndex weight that has results.
85%
85%
SpeedOutput tokens per second.
—
62/s
EloLMArena rating and rank.
—
1413
ReleasedMonth of release.
2026-02
2025-09
Input / 1MUS dollars per 1M input tokens.
$0.60 best
$0.78
Output / 1MUS dollars per 1M output tokens.
$3.60 best
$3.90
1K chats2K tokens in, 500 out, each.
$3.00 best
$3.51
1K documents20K tokens in, 1K out, each.
$15.60 best
$19.50
1K agent tasks200K tokens in, 20K out, each.
$192.00 best
$234.00
ContextMaximum tokens in one request.
128K
262K best
AgenticMulti-step tasks with tools.
59.4 best
47.1
CodingCode writing and repair.
54.2 best
46.3
ReasoningLogic problems and puzzles.
50.2 best
40.3
MultimodalTasks with images and text.
55.8
—
KnowledgeFacts and expert knowledge.
56.3 best
48.1
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
62.7 best
30.2
MathMath problems.
—
46.3
Benchmarks
6 sharedOnly Qwen3.5 397B (Reasoning): 3Only Qwen3 Max: 28
Qwen3.5 397B (Reasoning)
Qwen3 Max