# Voyage Code 2 Voyage Code 2 is Voyage AI by MongoDB's code-specialized embedding model. It improves code retrieval by 14.52% over OpenAI text-embedding-3-large across 11 datasets, with a context window of 0 tokens covering Python, C++, Java, and major ML frameworks. - **Model ID:** `voyage/voyage-code-2` - **Type:** embedding - **Providers:** voyage - **Context window:** Not applicable - **Maximum output tokens:** 0 - **Pricing:** $0.12/1M input tokens - **Canonical page:** https://vercel.com/ai-gateway/models/voyage-code-2 ## Supported parameters Detailed capability metadata has not been reported for this model. ## Example ```ts import { embed } from 'ai'; const result = await embed({ model: 'voyage/voyage-code-2', value: 'Sunny day at the beach', }) ``` ## About Voyage Code 2 is Voyage AI by MongoDB's code-specialized embedding model, released January 1, 2024. It features a context window of 0 tokens and targets code retrieval, code completion, and code assistant applications. On code retrieval tasks across 11 datasets derived from HumanEval, APPS, MBPP, DS-1000, CodeChef, and LeetCode, Voyage Code 2 achieves a 14.52% improvement in recall@5 over OpenAI text-embedding-3-large. Voyage Code 2 also performs well on general-purpose text retrieval, exceeding OpenAI text-embedding-3-large by 3.03% and Cohere Embed v3 by 4.93%. You can use a single embedding model for both code and documentation retrieval rather than maintaining separate indices with different models. Voyage AI by MongoDB evaluates it on Python, C++, and Java, plus documentation and usage patterns for Matplotlib, NumPy, Pandas, PyTorch, SciPy, scikit-learn, and TensorFlow. The model handles both natural language queries searching for code (text-to-code) and code snippets searching for similar code (code-to-code). ## What to consider Voyage Code 2 targets code search. If you're embedding source code, function signatures, and documentation for retrieval, it outperforms general-purpose embedding models by a wide margin. Voyage AI by MongoDB released voyage-code-3, which supports 300+ programming languages, a 32K context window, and Matryoshka dimensionality. Use voyage-code-3 for new deployments unless you need compatibility with existing Voyage Code 2 indices. Despite its code focus, Voyage Code 2 outperforms several general-purpose models on standard text retrieval. Use it for mixed code-and-documentation corpora without a second model. ## When to use ### Best For - **Code search engines** that retrieve relevant functions, classes, or modules from natural language queries - **Code completion pipelines** that use retrieval-augmented generation to find similar code patterns - **Developer documentation search** over API references, library docs, and code examples - **Mixed code and text retrieval** where a single model handles both source code and natural language documentation - **ML framework documentation** retrieval for Python-centric data science and machine learning workflows ### Consider Alternatives When - **You need broader language coverage** beyond Python, C++, and Java; voyage-code-3 supports 300+ programming languages - **You need a longer context window**; voyage-code-3 offers 32K tokens versus Voyage Code 2's 0 tokens - **Your workload is general-purpose text** with no code; a general-purpose embedding model like voyage-3.5 fits better - **You need Matryoshka dimensionality** for flexible embedding sizes; voyage-code-3 supports 2048/1024/512/256 dimensions ## Best for - **Code search engines:** Retrieve relevant functions, classes, or modules from natural language queries - **Code completion pipelines:** Retrieval-augmented generation finds similar code patterns - **Developer documentation search:** API references, library docs, and code examples - **Mixed code and text retrieval:** A single model handles both source code and natural language documentation - **ML framework documentation:** Retrieval for Python-centric data science and machine learning workflows ## Consider alternatives - **You need broader language coverage:** voyage-code-3 supports 300+ programming languages beyond Python, C++, and Java - **You need a longer context window:** voyage-code-3 offers 32K tokens versus Voyage Code 2's 0 tokens - **Your workload is general-purpose text with no code:** A general-purpose embedding model like voyage-3.5 fits better - **You need Matryoshka dimensionality:** voyage-code-3 supports 2048/1024/512/256 dimensions for flexible sizing ## Frequently asked questions ### What programming languages does Voyage Code 2 support? Voyage Code 2 is evaluated on Python, C++, and Java, along with documentation for major ML libraries (NumPy, Pandas, PyTorch, scikit-learn, TensorFlow, Matplotlib, SciPy). It handles both text-to-code and code-to-code retrieval. ### How does Voyage Code 2 compare to OpenAI text-embedding-3-large on code? Voyage Code 2 achieves a 14.52% improvement in recall@5 over OpenAI text-embedding-3-large across 11 code retrieval datasets derived from HumanEval, APPS, MBPP, DS-1000, CodeChef, and LeetCode. ### Should I use Voyage Code 2 or voyage-code-3? For new deployments, voyage-code-3 is the recommended choice. It supports 300+ programming languages (vs. Python, C++, Java), a 32K context window (vs. 16K), and flexible Matryoshka dimensions. Use Voyage Code 2 if you have existing indices and want to avoid re-embedding. ### Can Voyage Code 2 handle general text retrieval too? Yes. Voyage Code 2 exceeds OpenAI text-embedding-3-large by 3.03% on general-purpose text retrieval, making it viable for mixed code-and-documentation corpora. ### What is the context window for Voyage Code 2? 0 tokens. This handles most individual source files and documentation pages. For larger code contexts, use voyage-code-3 with its 32K-token window. ### How do I authenticate Voyage Code 2 through Vercel AI Gateway? Add your Voyage AI by MongoDB API key in AI Gateway settings, then send embedding requests through AI Gateway. AI Gateway authenticates requests across providers. ### What retrieval tasks is Voyage Code 2 evaluated on? 11 code retrieval datasets totaling 43,909 query-document pairs, derived from HumanEval, APPS, MBPP, DS-1000, CodeChef, and LeetCode. The evaluation metric is `recall@5`. ## Links - [Model page](https://vercel.com/ai-gateway/models/voyage-code-2) - [AI Gateway documentation](https://vercel.com/docs/ai-gateway) - [Provider model documentation](https://docs.voyageai.com/docs/embeddings) - [Provider pricing](https://docs.voyageai.com/docs/pricing)