DeepSeek V3.2 is a large language model designed to harmonize high computational efficiency with strong reasoning and agentic tool-use performance. Running inside a TEE (Trusted Execution Environment), with provider attestation support.
Context Window
164.0K
Max Output
65.5K
Input Price (Auto)
$0.50/1M
Output Price (Auto)
$1.00/1M
Performance metrics and benchmarks
Sourced from Artificial Analysis.
Intelligence Index
28.4
Auto routing is available for this model. Explicit provider selection is not available.
Loading provider options…
Coding Index
30.0
GPQA Diamond
Graduate-level scientific reasoning
73.8%
Better than 69% of models compared
HLE
Humanity's Last Exam
8.6%
Better than 67% of models compared
IFBench
Instruction-following benchmark
43.1%
Better than 55% 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
43.0%
Better than 60% of models compared
SciCode
Python programming for scientific computing
39.9%
Better than 83% of models compared
Terminal-Bench Hard
Agentic coding and terminal use
25.0%
Better than 74% of models compared
AIME 2025
American Invitational Mathematics Examination 2025
57.7%
Better than 55% of models compared
MMLU-Pro
Professional and academic subject knowledge
83.6%
Better than 85% of models compared
Last updated May 15, 2026
Artificial AnalysisLiveCodeBench
Contamination-free coding benchmark
55.4%
Better than 62% of models compared
