Mistral, a French AI lab, has officially launched its latest model, Large 4, also known as Le Chonk, which is now classified as an open weight model. This release marks a significant step forward in the competitive landscape of AI, particularly in the coding and cybersecurity sectors. The announcement comes with a set of impressive performance metrics that highlight the model's capabilities, positioning it as a frontrunner in the open model arena.

According to Mistral, Le Chonk has achieved remarkable Artificial Analysis scores, which are critical for evaluating AI performance in coding tasks. This achievement has led to claims that Le Chonk is the top open model available for coding, surpassing previous benchmarks set by other models in the field. The release of the weights for Large 4 is expected to enhance its accessibility and usability among developers and researchers alike, fostering innovation in AI applications.S2

In a bold assertion, Mistral states that Le Chonk has leapfrogged Chinese AI models in the open-weight race, claiming that it matches or even outperforms leading models from competitors like DeepSeek and Z.ai. This claim underscores Mistral's ambition to position itself at the forefront of AI development, particularly as competition intensifies globally. The implications of this advancement could reshape the landscape of AI tools available for various applications, including cybersecurity and software development.S3