Deepseek V3.2 Exp Thinking, Deepseek's latest model offering far better performance especially on longer contexts than its predecessors. Current flagship model by Deepseek. FP8.
Context Window
163.8K
Max Output
65.5K
Input Price (Auto)
$0.28/1M
Output Price (Auto)
$0.42/1M
Capabilities
Performance metrics and benchmarks
Sourced from Artificial Analysis.
Intelligence Index
32.9
Auto routing is available for this model. Explicit provider selection is not available.
Loading provider options…
Coding Index
33.3
Agentic Index
28.7
GPQA Diamond
Graduate-level scientific reasoning
79.7%
Better than 82% of models compared
HLE
Humanity's Last Exam
13.8%
Better than 82% of models compared
IFBench
Instruction-following benchmark
54.1%
Better than 73% of models compared
T²-Bench Telecom
Conversational AI agents in dual-control scenarios
33.9%
Better than 48% of models compared
AA-LCR
Long context reasoning evaluation
69.0%
Better than 94% of models compared
GDPval-AA
Economically valuable tasks
25.2%
Better than 83% of models compared
CritPt
Research-level physics reasoning
1.4%
Better than 88% of models compared
SciCode
Python programming for scientific computing
37.7%
Better than 75% of models compared
Terminal-Bench Hard
Agentic coding and terminal use
31.1%
Better than 80% of models compared
AIME 2025
American Invitational Mathematics Examination 2025
87.7%
Better than 85% of models compared
MMLU-Pro
Professional and academic subject knowledge
85.0%
Better than 91% of models compared
AA-Omniscience Accuracy
Proportion of correctly answered questions
28.2%
Better than 93% of models compared
Last updated May 15, 2026
Artificial AnalysisLiveCodeBench
Contamination-free coding benchmark
78.9%
Better than 91% of models compared
AA-Omniscience Hallucination Rate
Rate of incorrect answers among non-correct responses
81.1%
Better than 55% of models compared
