DeepSeek, operating under the username deepseek-ai, maintains a substantial public presence on GitHub. The organization hosts a wide range of repositories primarily using Python, C++, Cuda, TypeScript, and Rust. Notable projects include DeepSeek-V3, DeepSeek Coder, and the DeepSeek Harness, which focus on various aspects of AI and software integration.
Top languages
Public repositories
deepseek-harness
DeepSeek Harness: Everything is a Plugin.
DeepSeek-V3
No description provided for this repository.
DeepSeek-R1
No description provided for this repository.
awesome-deepseek-integration
Integrate the DeepSeek API into popular software
DeepSeek-Coder
DeepSeek Coder: Let the Code Write Itself
DeepSeek-OCR
Contexts Optical Compression
DeepInfer
DeepInfer: High-performance, large-scale inference framework for DeepSeek models
Janus
Janus-Series: Unified Multimodal Understanding and Generation Models
FlashMLA
FlashMLA: Efficient Multi-head Latent Attention Kernels
3FS
A high-performance distributed file system designed to address the challenges of AI training and inference workloads.
DeepEP
DeepEP: an efficient expert-parallel communication library
DeepGEMM
DeepGEMM: clean and efficient BLAS kernel library on GPU
open-infra-index
Production-tested AI infrastructure tools for efficient AGI development and community-driven innovation
DeepSeek-LLM
DeepSeek LLM: Let there be answers
DeepSpec
DeepSpec: a full-stack codebase for training and evaluating speculative decoding algorithms
DeepSeek-Coder-V2
DeepSeek-Coder-V2: Breaking the Barrier of Closed-Source Models in Code Intelligence
awesome-deepseek-agent
No description provided for this repository.
DeepSeek-VL2
DeepSeek-VL2: Mixture-of-Experts Vision-Language Models for Advanced Multimodal Understanding
DeepSeek-V2
DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model
smallpond
A lightweight data processing framework built on DuckDB and 3FS.
Engram
Conditional Memory via Scalable Lookup: A New Axis of Sparsity for Large Language Models
DeepSeek-VL
DeepSeek-VL: Towards Real-World Vision-Language Understanding
DeepSeek-OCR-2
Visual Causal Flow
DeepSeek-Math
DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
DreamCraft3D
[ICLR 2024] Official implementation of DreamCraft3D: Hierarchical 3D Generation with Bootstrapped Diffusion Prior
DualPipe
A bidirectional pipeline parallelism algorithm for computation-communication overlap in DeepSeek V3/R1 training.
deepseek-recipe
No description provided for this repository.
TileKernels
A kernel library written in tilelang
DeepSeek-MoE
DeepSeekMoE: Towards Ultimate Expert Specialization in Mixture-of-Experts Language Models
DeepSeek-V3.2-Exp
No description provided for this repository.
DeepSeek-Math-V2
No description provided for this repository.
EPLB
Expert Parallelism Load Balancer
DeepSeek-Prover-V2
No description provided for this repository.
profile-data
Analyze computation-communication overlap in V3/R1.
awesome-deepseek-coder
A curated list of open-source projects related to DeepSeek Coder
ESFT
Expert Specialized Fine-Tuning
DeepGEMM-Ascend
DeepGEMM-Ascend: clean and efficient matrix multiplication kernel library for Huawei Ascend NPUs
DeepSeek-Prover-V1.5
No description provided for this repository.
LPLB
An early research stage expert-parallel load balancer for MoE models based on linear programming.
DeepSelect
DeepSelect: TopK kernels for DeepSeek Sparse Attention (DSA) and Samplers
DeepJIT
A lightweight library for xPU kernel JIT compilation
DeepEP-Ascend
A high-performance communication library for machine learning training and inference on Huawei Ascend NPUs.
dsh-libreoffice-kit
An internal component used by DeepSeek Harness
clangd-ascend
Clangd for Ascend for Code Lint & Code Completion
dsh-node-addon-require-builtin
An internal component used by DeepSeek Harness
Frequently asked questions
What does deepseek-ai build on GitHub?
Deepseek-ai builds a variety of projects on GitHub, with repositories focusing on artificial intelligence, software integration, and automation. Their notable projects include DeepSeek-V3 and DeepSeek Coder, which showcase their innovative approaches to coding and AI applications.
Which programming languages does deepseek-ai use?
Deepseek-ai utilizes several programming languages in their repositories, including Python, C++, Cuda, TypeScript, and Rust. This diverse language usage reflects their multifaceted approach to software development and AI solutions.
Are deepseek-ai's repositories public?
Yes, all of deepseek-ai's repositories are public on GitHub. This openness allows other developers and researchers to explore, contribute to, and utilize their innovative projects, fostering collaboration within the AI community.
Is this exposure intended?
Monitor DeepSeek with RepoGuard and get alerted the moment a new public repository appears.
Monitor this account