IN

Inkling-Small

thinkingmachines

open weights·Active·multimodal general·Updated Jul 27, 2026

Inkling-Small is a 266.0B-parameter Mixture-of-Experts (MoE) image text to text model developed by thinkingmachines. Built on the InklingForConditionalGeneration architecture using transformers. Released on 2026-07-27 with 160 likes and 840 downloads on Hugging Face.

Inkling-Small

Model Overview

Inkling-Small is a 266.0B-parameter Mixture-of-Experts image text to text model developed by thinkingmachines. Built on the InklingForConditionalGeneration architecture. Released on 2026-07-27.


📊 Quick Specs

Specification Table
SpecificationValue
Parameters266.0B
ArchitectureInklingForConditionalGeneration
Taskimage text to text
Modalitytext, image
LicenseAPACHE-2.0
Frameworktransformers
MoEYes
Languages

✨ Key Features

  • 266.0B parameters with sparse MoE architecture for efficient inference
  • Built on InklingForConditionalGeneration architecture (transformers)
  • Primary task: image text to text (text, image modality)
  • Open-weights under APACHE-2.0 license — self-hostable and fine-tunable

📈 Community Adoption

  • 160 likes on Hugging Face
  • 840 downloads on Hugging Face

🔗 Resources


📜 License & Access

APACHE-2.0 — Open-weights model available for download, fine-tuning, and self-hosted deployment.

Related models comparison