
Thinking Machines releases open Inkling 2 Mini, a looped transformer model
Thinking Machines released Inkling 2 Mini, an open looped transformer with 26.4 billion parameters and 5.4 billion active per token. It scores 48.2 percent on DeepSWE v1.1 and runs locally on Macs with at least 32 GB of memory.
Thinking Machines has released Inkling 2 Mini, an open mixture-of-experts model built as a looped transformer. The model runs 12 of its 24 layers three times in a row, allowing internal reasoning without generating additional text. It is licensed under Apache 2.0 and supports text and images with a 1-million-token context window.
Looped transformer design
A looped transformer runs certain layers multiple times for each token. Inkling 2 Mini has 26.4 billion parameters, with 5.4 billion active per token because of the loops, according to Thinking Machines. The company says this lets the model reason internally for longer without generating more text, finding the right answer with fewer tokens.
Benchmarks and efficiency
Inkling 2 Mini leads all open models under 100 billion parameters on agentic coding tests, scoring 48.2 percent on DeepSWE v1.1 and 46.5 percent on the Vals Index v2.2. Artificial Analysis gives it 20 points on its Intelligence Index, tying it with Qwen3.8 27B at roughly six times lower cost and four times the speed. It also ties with Qwen3.8 27B (xhigh) and Nanbeige4.3-33B on that index, and with far fewer active parameters it sits directly on the Pareto frontier.
Availability
Inkling 2 Mini runs locally on Macs with at least 32 GB of memory and on Nvidia's DGX Spark. Weights and a research version with variable loops are available from Hugging Face, and the model can be tried in the Tinker Playground. Earlier coverage listed Inkling 2 Small on November 6, 2026 and Inkling Small on July 31, 2026.
Sources: The Decoder
SiTech — AI-powered web development
We build fast, modern websites and bring AI into real business workflows. Have a project or a question? We'd love to help.