Mistral has debuted Pixtral 12B, a multimodal AI model capable of processing both images and text, available for free for research and academic use.
Pixtral 12B is built on Mistral's previous text model, Nemo 12B, allowing it to answer queries about images of varying size when provided via URLs or base64.
While the model promises to perform tasks like image captioning and object counting, no operational demos were available at publication, creating uncertainty for users.
Mistral recently completed a $645 million funding round, leading to a valuation of $6 billion as it expands its AI capabilities with the Pixtral launch.
[
Collection
]
[
|
...
]