AI

Mistral claims its newest AI model delivers leading performance for the price

French AI Startup Mistral releases a new AI model, Mistral Medium 3, which is aimed at efficiency without jeopardizing performance.

Available in the API of Mistral priced at $ 0.40 per million input tokens and $ 2 per million output tokens, Mistral Medium 3 performs “AT of about” 90% of Cleptlier Claude Sonnet 3.7 model from Anthropic on “Benchmarks over the board”, says Mistral. It also surpasses recent open models, including Meta Llama 4 Maverick and Cohere’s Command A on popular AI performance evaluations.

Tokens are the rough pieces of data models, with a million tokens that are equivalent to around 750,000 words (about 163,000 words longer than “war and peace”).

“Mistral Medium 3 […] Be deployed on every cloud, including self -hosted environments of four GPUs and higher, “Mistral explained in a blog post that was sent to WAN.” On prices, the modeling costs leaders such as Deepseek V3, both in API and self -written systems. “

Mistral Medium 3
The performance of Mistral Medium 3 on AI benchmarks.Image Credits:Mistral

Mistral, founded in 2023, is a Laboratory for Frontier, with the aim of building a series of AI-driven services, including a chatbot platform, Le Chat and mobile apps. It is supported by VCs including a general catalyst and has so far raised more than € 1.1 billion (around $ 1.24 billion). Mistral’s customers are BNP Paribas, Axa and Mirakl.

According to Mistral, Mistral Medium 3 is the best for coding and stem tasks and excels in multimodal concept. The company says that customers in financial services, energy and health care tests that test the model for use cases such as customer service, workflow automation and analyzing complex data sets.

See also  The AI Price War: How Lower Costs Are Making AI More Accessible

In addition to the API of Mistral, where Enterprise customers can work with Mistral to refine it, Milrtral Medium 3 is available on the Amazon Sagemaker platform from Wednesday. It quickly comes to other hosts, including Microsoft’s Azure Ai Foundry and Google’s Vertex AI platforms, the company added.

WAN event

Berkeley, Ca
|
June 5

Book now

The launch of Mistral Medium 3 follows the heels of Mistral’s Mistral Small 3.1 in March. In his blog post, the company teased the release of a much larger model in the coming weeks.

Mistral also launched Le Chat Enterprise on Wednesday, a business -oriented chatbot service that offers tools such as an AI “agent” builder and integrates the models of Mistral with external services such as Gmail, Google Drive and SharePoint. Le Chat Enterprise was rolled out in private preview earlier this year, but today marks the general availability.

Le Chat Enterprise will soon support MCP, the standard of Anthropic for connecting AI assistants with the systems and software where there is data. Other large AI model providers, including Google and OpenAi, have announced that they would adopt MCP earlier this year.

Updated 7:48 AM Pacific: A price errors resolved – the version of Mistral’s Blogpost WAN received a typo. We regret the mistake

Source link

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button