- Mistral Small: Perfect for edge deployments, this lightweight version has quicker inference.
- OpenOrca: Instruction-tuned for complicated task following and reasoning.
- Mixtral: Mixtral's vision-language counterpart for multimodal inputs (text + picture).
· Mistral Nemo: Depending on the dataset, it often focuses on conversation or high-quality summarization.
