It’s been two months since Voyage AI joined the MongoDB team, so I wanted to update everyone on how things are progressing. All of us here at Voyage AI remain focused on our core mission: developing state-of-the-art embedding models and rerankers. 

And I want to state explicitly that Voyage AI models are—and will continue to be—available for all teams, whether you are using MongoDB or not.

Even though it’s only been a few months, being part of MongoDB is already rapidly accelerating Voyage AI’s work. With expanded investment in Engineering and Research (we’re actively hiring – check out our open roles!) and increased compute resources, Voyage AI can support more development teams with the best models in the industry. One team, Continue, recently wrote about successfully using Voyage AI models to improve their leading open-source AI code assistant.

A few exciting developments to keep an eye on in the near term:

• Scaling the Voyage AI APIs: We are building the Voyage AI platform to handle 100x traffic to offer enhanced reliability and reduced latency.
• More cutting-edge models: Our research team is working hard to introduce new models (with more capabilities) that will give development teams additional high-quality model options for developing AI applications.
• Automated text embedding in MongoDB Atlas Vector Search: We’re building a one-click semantic search experience with automatic embedding generation at ingestion, update, and query—no external models or pipelines needed.
• MongoDB Atlas Integration: For existing MongoDB customers, direct access to Voyage AI model APIs through MongoDB Atlas will make it easy to adopt Voyage AI’s suite of models.

Everyone at Voyage AI is excited for all we have in store for developers in the coming months.

Tengyu Ma

Wondering how to stay connected with us?

• Come say hi to the team in San Francisco at the MongoDB booth at LangChain Interrupt next week, or join one of our sessions at the AI Engineers World’s Fair in early June!
• Explore the models in our documentation.
• Register your interest for the private preview of automated text embeddings in MongoDB Atlas Vector Search, which will be available later this year.
• Follow our blog and social channels on X and LinkedIn to stay updated on our research, model releases, and integration with MongoDB Atlas.