Inkling by Thinking Machines is here, a 1 TRILLION parameter open model that natively understands images, text, AND audio, with a 1M token context window. In this video, we break down the architecture, show you how to run it, and share our vibe-eval results. Now on Hugging Face 🤗 https://huggingface.co/thinkingmachines/Inkling ⚡ TL;DR • 975B total / 41B active params (MoE, 256 experts) • Native image, text & audio input — one model, no separate encoders • 1M context window, trained on 45T tokens • BF16 + calibrated NVFP4 checkpoints, MTP layers for speculative decoding • Day-0 support: transformers, SGLang, vLLM, llama.cpp ⏱️ CHAPTERS 00:00 Intro — What is Inkling? 05:50 Architecture deep dive (Relative attention, Hybrid attention, SConv) 10:00 Interview with Thinking Machines (John Schulman & Soumith Chintala) 13:12 Vision & audio towers explained 16:18 Why 1 Trillion parameters? 18:22 Relative embedding vs. Rotary (RoPE) 21:28 Working with PyTorch modules in Transformers 22:11 1-bit GGUFs with llama.cpp & Unsloth 25:07 Hardware requirements & performance (8x H100 benchmarks) 27:38 Running it: Transformers pipeline & source code 29:21 Serving with SGLang & vLLM 29:53 Agentic coding demo with Pi harness 31:09 Vision & audio vibe evals (Physics puzzles & Big Bench) 33:10 Post-training with Tinker + OpenEnv 34:38 Free inference via Inference Providers 35:20 Final thoughts and credits 🔧 TRY IT • Model (BF16): https://huggingface.co/thinkingmachines/Inkling • Model (NVFP4): https://huggingface.co/thinkingmachines/Inkling-NVFP4 • GGUF quants: https://huggingface.co/unsloth/inkling-GGUF • Inference Providers (free for 2h at launch): https://huggingface.co/thinkingmachines/inkling?inference_api=true • Quick start: pip install -U transformers, then pipeline("any-to-any", model="thinkingmachines/Inkling") 📚 RESOURCES • Full blog post: https://huggingface.co/blog/thinkingmachines-inkling • Vibe eval images & results: https://huggingface.co/buckets/merve/inkling • RL example (tinker + OpenEnv): https://github.com/huggingface/OpenEnv • Distillation with TRL (GOLD): https://github.com/huggingface/trl 💡 KEY TAKEAWAYS • Needs ~2TB VRAM in BF16 (600GB in NVFP4) — but 1-bit GGUFs retain ~74% top-1 accuracy at 86% smaller • Reasoning effort is tunable from "none" to "max"; medium (0.7) is the sweet spot • The model transcribes/OCRs inputs first, then reasons — prompt accordingly • Great for multimodal reasoning apps, document processing, and fine-tuning 👥 By burtenshaw, merve, pcuenq & ariG23498 from Hugging Face #Inkling #ThinkingMachines #HuggingFace #OpenSource #LLM #Multimodal #AI #MachineLearning