French AI startup Mistral is set to release a new AI model, Mistral Medium 3, which it claims offers top-tier performance at an affordable price. Available through Mistral's API at $0.40 per million input tokens and $20.80 per million output tokens, the company asserts that Mistral Medium 3 matches or exceeds the performance of Anthropic’s more expensive Claude Sonnet 3.7 model across various benchmarks. Moreover, it outperforms recent open-source models, such as Meta’s Llama 4 Maverick and Cohere’s Command A, in popular AI evaluations. To put this into perspective, tokens are the fundamental units of data that AI models process. One million tokens roughly equate to 750,000 words, which is approximately 163,000 words longer than Leo Tolstoy's "War and Peace." Mistral emphasizes that Medium 3 can be deployed on any cloud environment, including self-hosted systems with four or more GPUs. “Notably, the model’s pricing is competitive, beating cost leaders like DeepSeek v3 in both API and self-deployed systems,” according to a blog post shared with TechCrunch. Founded in 2023, Mistral aims to develop a variety of AI-powered services, such as a chatbot platform called Le Chat and mobile applications. The company has attracted significant investment, raising over €1.1 billion (approximately $1.24 billion) from venture capitalists, including General Catalyst. Prominent clients like BNP Paribas, AXA, and Mirakl have already begun beta testing Mistral Medium 3 for applications ranging from customer service and workflow automation to analyzing complex data sets in sectors like financial services, energy, and healthcare. Mistral Medium 3 excels particularly in coding and STEM-related tasks, as well as multimodal understanding, which involves processing and interpreting multiple types of data, such as text and images. Enterprise customers can access the model through Mistral’s API, where they can collaborate with the company to fine-tune it for specific use cases. Additionally, Mistral Medium 3 will be available on Amazon’s Sagemaker platform starting Wednesday, with plans to expand to Microsoft’s Azure AI Foundry and Google’s Vertex AI platforms in the near future. The introduction of Mistral Medium 3 comes shortly after the release of Mistral Small 3.1 in March. In their blog post, Mistral hinted at the upcoming launch of an even larger model, which is expected to further solidify their position in the AI market. This strategic rollout underscores the company’s commitment to providing high-performance, cost-effective AI solutions to a wide array of industries, making it a compelling player in the rapidly evolving field of artificial intelligence.