DeepSeek-V3.1

Hybrid AI model supporting both thinking and non-thinking modes

Visit DeepSeek-V3.1 →

DeepSeek-V3.1 is a large language model developed by DeepSeek that operates in both reasoning (thinking) and standard (non-thinking) modes, making it flexible for complex problem-solving and everyday tasks. It is designed for developers, researchers, and enterprises needing a capable frontier model with agentic capabilities. Notable for its strong benchmark performance and open-weight availability, it positions itself as a step toward autonomous AI agents.

At a glance

Company
DeepSeek
Pricing
freemium
API available
Yes
Self-hostable
Yes
Launched
2025-08
Last verified
2026-05-11

Capabilities

long-contextchain-of-thoughtagentic-reasoningmulti-modal-inputcode-generationfunction-calling

Categories

Alternatives

For AI agents: machine-readable markdown version of this page at /tools/deepseek-v3-1.md, or send Accept: text/markdown.