Mistral launches new code embedding model that outperforms OpenAI and Cohere in real-world retrieval tasks

Mistral AI has launched Codestral Embed, its new embedding model tailored for code retrieval and other tasks like semantic search and code analytics. The model outperforms existing offerings on benchmarks while offering flexibility in dimensions and precision levels. Available at $0.15 per million tokens, it stands out as a cost-effective solution with impressive performance, competing against models from major players like Cohere and OpenAI, and even open-source alternatives. Codestral Embed's versatility across multiple use cases makes it a significant tool for developers and enterprises seeking efficient code intelligence solutions.