0

Qwen3.6 Plus

Closed

Qwen3.6Plus is an AI model from Alibaba.

Publisher
Alibaba
Family
Qwen3.6
Context
1M
Input
Output
License
proprietary
Released
2 Apr 2026

Cite

Notes

Only stored in your browser.

Attribution

Benchmark scores
frontiermathAALiveBenchprime-hubOpenRewardVals AI
Attribution policy →
Context
1M
Input
$0.50
/1M tok
Output
$3
/1M tok
Blended
$1.13
#315/ 414/1M · 3:1
Speed
53
#174/ 414tok/s
Cache write
$0.41
0.81x input

Price & speed via Artificial Analysis · ranked across tracked models · caching via OpenRouter

Reported on 31 evals across 11 domains - best 97.7% on τ²-bench (Tau²-bench) · #5 of 320

Reported eval scores

31
MMLU Pro88.5
Infraresolutionbench83.2
SWE Bench Multilingual73.8
MMLU Redux 294.5
HMMT96.7
MMMLU89.5
AIME202695.3
τ²-bench (Tau²-bench)97.7
GPQA Diamond88.2
SciCode40.7
Terminal-Bench (Hard)43.9
IFBench75.2
LiveBench - Coding78.2
LiveBench - Language75
LiveBench - Data Analysis69.9
LiveBench - Math83.7
LiveBench - Reasoning75.8
LiveBench - Instruction Following58.3
Humanity's Last Exam (HLE)25.7
LiveBench73.5
Vals Index48
TaxEval v274.7
MortgageTax68
CorpFin v261.9
Finance Agent (v2)40.8
FrontierMath26.2
Vibe Code Bench v1.125.6
SAGE44.9
MedScribe77
MedCode36.9
Vals Multimodal Index50.7