Research

Enhance AI coding with state-
of-the-art code embedding

Qodo-Embed-1 is a family of large-scale embedding models to significantly enhance context retrieval for code-related tasks.

Code embedding

AI-powered coding isn’t just about generating new code—it’s about understanding, retrieving, and working with existing code efficiently. Code embeddings transform complex codebases into structured, searchable representations, enabling AI to find relevant snippets, improve context awareness, and power smarter developer tools.

The Qodo-Embed-1 Model Family

Qodo-Embed-1-1.5B

1.5B Parameters
Open weights

Qodo-Embed-1-7B

7B Parameters
Commercial License

Mascot

Enable code search and retrieval

Find and retrieve relevant or similar code to queries from large repositories

Enhance contextual awareness

Improve AI-generated code accuracy with relevant context

Natural language-to-code & code-to-code understanding

Convert natural language into code and match similar logic across language

Qodo-Embed outperforms leading code embedding models.

Benchmarked on COIR benchmark, designed for retrieval tasks.