240 likesNew · 12d old
Inkling 2 Small: Open-weight multimodal model for text-and-image prompts
What is Inkling-2-Small?
Inkling 2 Small is a multimodal AI that accepts both images and text and returns written answers. It works in English and other languages and can be fine-tuned for coding assistants, chatbots, or any app that needs visual-language understanding.
Who it is for
Developers building AI assistants or retrieval-augmented tools can use it to add image-aware responses.
How to try it
Open the Tinker Playground to test Inkling 2 Small.
The story so far
Early Signal first spotted thinkingmachines/Inkling-2-Small on November 13, 2026, with 240 likes. On its busiest day it gained 24 likes. The project is 12 days old. Early Signal files it under Coding, Image, Agents.
Questions about Inkling-2-Small
- How do I try Inkling-2-Small?
- Open the Tinker Playground to test Inkling 2 Small.
- What do I need to run Inkling-2-Small?
- GPU cluster with at least 500 GB aggregated VRAM (e.g., 2 × NVIDIA B300)