Private AI
Qwen3.5 35B A3B with extended reasoning enabled. A native vision-language MoE model with hybrid attention.
Added Feb 24, 2026
Model weightsContext Window
260.1K
Max Output
65.5K
Input Price (Auto)
$0.24/1M
Output Price (Auto)
$1.89/1M
Cache Read (Auto)
$0.12/1M
Capabilities
Performance metrics and benchmarks
Sourced from Artificial Analysis.
GPQA Diamond
Graduate-level scientific reasoning
84.5%
Better than 81% of models compared
Choose explicit providers for this model. Auto routing remains available as the default option.
Loading provider options…
HLE
Humanity's Last Exam
19.7%
Better than 78% of models compared
IFBench
Instruction-following benchmark
72.5%
Better than 89% of models compared
T²-Bench Telecom
Conversational AI agents in dual-control scenarios
89.2%
Better than 84% of models compared
AA-LCR
Long context reasoning evaluation
62.7%
Better than 75% of models compared
CritPt
Research-level physics reasoning
0.9%
SciCode
Python programming for scientific computing
37.7%
Better than 63% of models compared
Terminal-Bench Hard
Agentic coding and terminal use
26.5%
Better than 67% of models compared
AA-Omniscience Accuracy
Proportion of correctly answered questions
20.4%
AA-Omniscience Hallucination Rate
Rate of incorrect answers among non-correct responses
84.0%
Last updated Jul 6, 2026, 3:53 PM
Artificial Analysis