Thinking Machines, the startup led by former OpenAI CTO Mira Murati, has released Inkling Small, an open-source multimodal model with 276 billion total parameters. It scores 40 on the Artificial Analysis Intelligence Index, just one point behind the original Inkling's 41, despite the larger model having 975 billion parameters. Inkling Small activates only 12 billion parameters per token versus 41 billion for its predecessor. On several benchmarks it outperforms the flagship, including 80.2 percent on SWE-bench Verified and 64.7 percent on Terminal Bench 2.1. It accepts text, image and audio inputs, supports a one-million-token context window, and is released under the permissive Apache 2.0 license. Weights are available on Hugging Face and can be fine-tuned through the company's Tinker API. The model still requires substantial GPU memory—at least 600 GB for the full checkpoint—so it targets enterprise servers rather than laptops. Launch pricing starts at $0.58 per million input tokens
Thinking Machines, the startup led by former OpenAI CTO Mira Murati, has released Inkling Small, an open-source multimodal model with 276 billion total parameters. It scores 40 on the Artificial Analysis Intelligence Index, just one point behind the original Inkling's 41, despite the larger model having 975 billion parameters. Inkling Small activates only 12 billion parameters per token versus 41 billion for its predecessor. On several benchmarks it outperforms the flagship, including 80.2 percent on SWE-bench Verified and 64.7 percent on Terminal Bench 2.1. It accepts text, image and audio inputs, supports a one-million-token context window, and is released under the permissive Apache 2.0 license. Weights are available on Hugging Face and can be fine-tuned through the company's Tinker API. The model still requires substantial GPU memory—at least 600 GB for the full checkpoint—so it targets enterprise servers rather than laptops. Launch pricing starts at $0.58 per million input tokens