DeepSeek-V3.2
#3τ-bench
by DeepSeek
+
+
+
+
About
DeepSeek-V3.2, released by DeepSeek on December 1, 2025, is a large language model with 685 billion total parameters featuring integrated thinking in tool-use and support for both reasoning and direct generation modes. It features a 128K token context window and introduced large-scale agent training across 1,800+ environments. DeepSeek-V3.2 targets agentic workflows, complex instruction following, and coding tasks under an open MIT license.
+
+
+
+
Pricing Range
Input (per 1M)$0.27 -$0.27
Output (per 1M)$1.10 -$1.10
Providers1
+
+
+
+
Timeline
ReleasedDec 1, 2025
+
+
+
+
License & Family
License
MIT
Performance Overview
Performance metrics and category breakdown
Overall Performance
2 benchmarks
Average Score
59.0%
Best Score
80.4%
High Performers (80%+)
1Performance Metrics
Max Context Window
136.0KTop Categories
Agents
80.4%
Coding
37.5%
+
+
+
+
All Benchmark Results for DeepSeek-V3.2
Complete list of benchmark scores with detailed information
| τ-bench | Agents | 80.40 | 80.4% | Unverified | |
| SWE-rebench | Coding | 37.50 | 37.5% | Unverified |