Voyage
An AI company specializing in embedding models. Its core products are vectorization models optimized for retrieval-augmented generation (RAG) scenarios, with targeted improvements in retrieval accuracy for technical documentation, code, and API reference content.
Core Facts
Timeline (last 90 days)
chatVia.ai supports OpenAI, Mistral, Gemini, Cohere, Jina, Voyage, Qwen3 or OpenRouter as embedding models, and stores vectors as halfvec with an HNSW index built.
chatVia.ai uses cross-encoder rerankers (Cohere, Jina, Voyage, etc.) to rerank the fused candidates, and sets a similarity gate to discard weak matches.
The only change in Voyage 2.0.69 is updating the dependency @ai-sdk/provider-utils to version 5.0.69.
The jump from 2.0.68 to 2.0.69 theoretically means callers can upgrade without modifying any code.
The Voyage 2.0.69 update introduces no new APIs, makes no changes to the public interface, and involves no configuration item changes.
The commit record for the Voyage 2.0.69 release shows only one commit, and it is GitHub-verified signed.
For projects using the Voyage provider, the direct benefit of upgrading to 2.0.69 is converging @ai-sdk/provider-utils in the dependency tree, avoiding TypeScript type conflicts caused by multiple versions coexisting.
Voyage is one of the providers focused on embedding models; @ai-sdk/voyage only needs to implement the embedding-related interfaces, making it smaller in size than language model adapter packages.
The public API of @ai-sdk/voyage did not change in version 1.0.38; the calling method, parameter structure and return format of Voyage embedding all remain the same.
Joey uses Voyage embeddings to prioritize retrieval of the most relevant documents in order to reduce search time.
3 more timeline events
All Facts (12)
Vercel AI SDK provides unified abstract interfaces (such as generateText and streamText), allowing developers to maintain a consistent coding style when calling different model providers.
90%UnverifiedchatVia.ai uses cross-encoder rerankers (Cohere, Jina, Voyage, etc.) to rerank the fused candidates, and sets a similarity gate to discard weak matches.
50%UnverifiedchatVia.ai supports OpenAI, Mistral, Gemini, Cohere, Jina, Voyage, Qwen3 or OpenRouter as embedding models, and stores vectors as halfvec with an HNSW index built.
50%UnverifiedThe Voyage 2.0.69 update introduces no new APIs, makes no changes to the public interface, and involves no configuration item changes.
50%UnverifiedThe commit record for the Voyage 2.0.69 release shows only one commit, and it is GitHub-verified signed.
50%UnverifiedThe only change in Voyage 2.0.69 is updating the dependency @ai-sdk/provider-utils to version 5.0.69.
50%UnverifiedFor projects using the Voyage provider, the direct benefit of upgrading to 2.0.69 is converging @ai-sdk/provider-utils in the dependency tree, avoiding TypeScript type conflicts caused by multiple versions coexisting.
50%UnverifiedVoyage is one of the providers focused on embedding models; @ai-sdk/voyage only needs to implement the embedding-related interfaces, making it smaller in size than language model adapter packages.
50%UnverifiedThe public API of @ai-sdk/voyage did not change in version 1.0.38; the calling method, parameter structure and return format of Voyage embedding all remain the same.
50%UnverifiedJoey uses Voyage embeddings to prioritize retrieval of the most relevant documents in order to reduce search time.
50%UnverifiedJoey uses Voyage's embedding model for document retrieval because Voyage has targeted optimizations for retrieval precision on technical documentation, code, and API reference content.
50%UnverifiedThe jump from 2.0.68 to 2.0.69 theoretically means callers can upgrade without modifying any code.
50%