Thinking Machines releases Inkling, a 1T-parameter multimodal LLM with 1M context

Thinking Machines has released Inkling, a large multimodal language model with approximately 1 trillion parameters and a 1 million token context window. The model natively processes text, image, and audio inputs, and features agentic capabilities. Inkling is built on a decoder-only Mixture-of-Experts architecture with hybrid attention mechanisms and is available with open weights for research and fine-tuning. AI

IMPACT Sets a new benchmark for multimodal LLMs with large context windows, potentially accelerating development of complex reasoning applications.

RANK_REASON Frontier-lab model release with system card [lever_c_demoted from frontier_release: ic=2 ai=1.0]

Read on Hugging Face Blog →

• Hugging Face
• Inkling
• Thinking Machines
• bfloat16
• GLM-5
• Kimi K2.5
• Kimi K2.6
• Mixture of Experts (MoE)
• Nemotron 3 Ultra
• NVFP4
• SGLang
• TokenSpeed
• Unsloth
• vLLM

• model release
• product

AI-generated summary · Google Gemini · from 2 sources. How we write summaries →