A massive open-weights multimodal AI model with a 1 million token context window.
Inkling AI is a large-scale open-weights multimodal model developed by Thinking Machines. With 975 billion parameters, it is designed to handle both text and visual inputs. A standout feature is its exceptionally large 1 million token context window, enabling it to process and reason over extensive documents or long-form content. Released under the permissive Apache 2.0 license, it is freely available for research and commercial use. The model is aimed at developers, researchers, and enterprises seeking a powerful, open foundation model for building advanced AI applications without restrictive licensing. It solves the need for a highly capable, transparent, and accessible large language model that can handle complex, long-context multimodal tasks.
-
975 billion parameter open-weights model
-
Multimodal (text and visual inputs)
-
1 million token context window
-
Apache 2.0 license for commercial use
-
Open benchmarks available for performance evaluation
-
Analyzing and summarizing long documents like legal contracts or research papers
-
Building advanced chatbots and virtual assistants that require extensive context memory
-
Developing AI applications for complex visual and textual reasoning tasks
-
Academic and industrial research on large language model capabilities
-
Creating content generation tools for long-form writing and multimedia