Qwen3.6 27B is a native vision-language dense model with stronger agentic coding and STEM reasoning than Qwen 3.5 27B. It also improves spatial intelligence (including object localization/detection), plus video understanding, document OCR, and visual-agent workflows.
Added Apr 23, 2026
Context Window
260.1K
Max Output
65.5K
Input Price
$0.60/1M
Output Price
$3.60/1M
Capabilities
Performance metrics and benchmarks
Sourced from Artificial Analysis.
Intelligence Index
37.1
Coding Index
26.6
GPQA Diamond
Graduate-level scientific reasoning
82.9%
Better than 85% of models compared
HLE
Humanity's Last Exam
13.6%
Better than 79% of models compared
IFBench
Instruction-following benchmark
45.7%
Better than 57% of models compared
T²-Bench Telecom
Conversational AI agents in dual-control scenarios
93.6%
Better than 92% of models compared
AA-LCR
Long context reasoning evaluation
55.0%
Better than 71% of models compared
SciCode
Python programming for scientific computing
37.3%
Better than 70% of models compared
Terminal-Bench Hard
Agentic coding and terminal use
21.2%
Better than 66% of models compared
Last updated May 11, 2026
Artificial Analysis