Designed for greater capability, faster inference, higher throughput, and scaling to larger models.
DeepSeek-V4-Pro is a frontier Mixture-of-Experts model with a large context window and three reasoning modes.
DeepSeek-OCR is a vision-language model that can perform token-efficient OCR.
DeepSeek-R1 is a family of open reasoning models with performance approaching that of leading models, such as O3 and Gemini 2.5 Pro.
A strong Mixture-of-Experts (MoE) language model with 671B total parameters with 37B activated for each token.
DeepSeek Coder is a capable coding model trained on two trillion code and natural language tokens.
An open-source Mixture-of-Experts code language model that achieves performance comparable to GPT4-Turbo in code-specific tasks.
DeepSeek-V3.1-Terminus is a hybrid model that supports both thinking mode and non-thinking mode.
A strong, economical, and efficient Mixture-of-Experts language model.
An advanced language model crafted with 2 trillion bilingual tokens.
An upgraded version of DeekSeek-V2 that integrates the general and coding abilities of both DeepSeek-V2-Chat and DeepSeek-Coder-V2-Instruct.
A fine-tuned version of Deepseek-R1-Distilled-Qwen-1.5B that surpasses the performance of OpenAI’s o1-preview with just 1.5B parameters on popular math evaluations.
A fully open-source family of reasoning models built using a dataset derived by distilling DeepSeek-R1.
A version of the DeepSeek-R1 model that has been post trained to provide unbiased, accurate, and factual information by Perplexity.
A reset version of Eriktonon's DeepSeekv4.1-Flash - Ready for use in AI Agent Scenarios
Omnigenesis Lyapunov stability specialist based on DeepSeek R1 1.5B
UNCENSORED (Abliterated + SecOps) — DeepSeek-V4.1-Flash-1017 (9B Compact & 27B Large 128K Edition) — Fine-tuned & Orthogonally Abliterated for Agents Architecture, Rust Systems & Autonomous Tool-Use.
1017