#6 · Primary category: Foundation Models
DeepSeek-Coder
DeepSeek Coder: Let the Code Write Itself
Project last updated:11/11/25
GitHub Stars
24.2K
Forks
2.9K
Contributors
17
License
MIT
Why we included this project
DeepSeek Coder is a family of code-specialized language models you can download and run yourself, which matters if you want a capable code model without depending on a hosted API. Sizes run from 1B to 33B, so you can match the model to your hardware and latency budget. Each one is trained on a project-level corpus with a 16K context window and a fill-in-the-blank objective, which is what lets the base models handle whole-file completion and infilling across dozens of languages. The instruction-tuned variants work as chat-style assistants for explaining or generating code. The repo ships Transformers examples for completion, insertion, and chat inference, plus a local demo app, so getting a model running is mostly pulling weights and writing a few lines. If you are building an internal coding tool or an IDE plugin and want full control over the model, this is a well-documented option worth evaluating.
Articles for this project
No articles for this project yet.
To suggest a topic or contribute an article, contact us.
Related projects in this category
transformers
🤗 Transformers: the model-definition framework for state-of-the-art machine learning models in text, vision, audio, and multimodal models, for both inference and training.
CLIP
CLIP (Contrastive Language-Image Pretraining), Predict the most relevant text snippet given an image
MiniCPM-V
A Pocket-Sized MLLM for Ultra-Efficient Image and Video Understanding on Your Phone
generative-models
Generative Models by Stability AI
unilm
Large-scale Self-supervised Pre-training Across Tasks, Languages, and Modalities