Inkling accepts both text and images as input, making it a multimodal model capable of working across visual and language tasks. Beyond its open-weight availability under the Apache 2.0 license, details about its specific strengths, limitations, and intended use cases are limited based on available information.