Welcome Inkling by Thinking Machines

Inkling now comes in smaller size 🤗 Inkling-Small is out by Thinking Machines Lab. We have updated this post with performance, and deployment configurations for the Inkling-Small and the Inkling-Small-NVFP4 variants. Here’s the collection with all the Inkling models. We made it easier for you to deploy Inkling-Small with one-click on Inference Endpoints (getting up to 160 TPS). We also ship a real-time voice and image demo where you can interact with the model.

Inkling is a large (1T params!) open model to natively accept image, text, and audio inputs.

TLDR; Inkling by Thinking Machines is out on Hugging Face. Inkling is a huge multimodal LLM that understands all modalities (image, audio, text), has agentic capabilities, and supports 1M context. It comes in full BF16 and a well-calibrated NVFP4 variant, and includes speculative MTP layers for faster inference. There’s day-0 support in transformers, SGLang, vLLM, and llama.cpp. Thinking Machine Labs also released Inkling-Small with 276B total and 12B active parameters, as well as support for MXFP8 and NVFP4 weights, with the same architecture as Inkling.

What makes Inkling special?

Inkling is the first large open model with ~1T parameters and 1M context window to natively receive image, text, and audio inputs, trained on 45 trillion tokens of text, images, audio and video. It’s focused on reasoning across modalities such as audio, images, and text; and is intended for domain adaptation via fine-tuning. We’ve tinkered with this model to build some demos and explore the architecture, and we think it’s great for building a new wave of multimodal reasoning apps.

Overall Capabilities and Architecture

Inkling is a decoder-only multimodal Mixture-of-Experts model with 975B total and 41B active parameters. There are a lot of things going on, so let’s break each part down:

Here’s a quick glance of the architecture.