dots3-note Preview vs Qwen3.8-Omni-Flash

Change models
dots3-note PreviewDots Studio
IndexAll categories, out of 100.
63.4 ±3.9 best
62.7 ±7.5
GapIndex points below the top.
Top
−0.7 · no clear gap
CoverageIndex weight that has results.
95% best
70%
SpeedOutput tokens per second.
—
41/s
EloLMArena rating and rank.
—
—
ReleasedMonth of release.
2026-08
2026-09
Input / 1MUS dollars per 1M input tokens.
Free best
$0.15
Output / 1MUS dollars per 1M output tokens.
Free best
$0.47
1K chats2K tokens in, 500 out, each.
Free best
$0.535
1K documents20K tokens in, 1K out, each.
Free best
$3.47
1K agent tasks200K tokens in, 20K out, each.
Free best
$39.40
ContextMaximum tokens in one request.
512K
1M best
AgenticMulti-step tasks with tools.
63.3
64.5 best
CodingCode writing and repair.
60.3
61.5 best
ReasoningLogic problems and puzzles.
74.5
—
MultimodalTasks with images and text.
59.2
64.5 best
KnowledgeFacts and expert knowledge.
77.2 best
60.8
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
58.2
65.2 best
MathMath problems.
62.3
—

Benchmarks

7 shared
Only dots3-note Preview: 20Only Qwen3.8-Omni-Flash: 11
dots3-note Preview
Qwen3.8-Omni-Flash
61.0%31 / 83
63.3% best21 / 83
52.6% best18 / 77
36.5%38 / 77
91.5%7 / 59
92.6% best3 / 59
80.4%11 / 57
81.5% best7 / 57
NL2RepoCoding
49.8% best12 / 36
48.9%13 / 36
MathVisionMultimodal
87.7%12 / 24
91.8% best5 / 24
83.1%11 / 17
83.5% best10 / 17