Inkling drops a 975B open-weights multimodal MoE ready for fine-tuning
Inkling is a 975‑billion‑parameter Mixture‑of‑Experts transformer with 41 billion active weights and a 1 million‑token context, supporting text, images, and audio. The full weights are released openly and can be fine‑tuned on the Tinker platform, giving developers a high‑capacity, multimodal foundation at open‑source cost, a rare alternative to closed‑source frontier models.