← Back
1 December 2025Model release
DeepSeek releases V3.2 and V3.2-Speciale
DeepSeek shipped MIT-licensed V3.2, its first model integrating thinking into tool use, and a high-compute V3.2-Speciale it claimed reached gold-level results at IMO and IOI 2025.
Actors involved
Tags
deepseekopen-weightsreasoning