New open model Inkling 2 Mini thinks in loops to solve problems with fewer words and still runs on a laptop
Summary
- Thinking Machines has released Inkling 2 Mini, an open mixture-of-experts looped transformer model with 26.4 billion total parameters and 5.4 billion active parameters per token.
- The model runs 12 of its 24 layers three times in a row, allowing it to reason internally and use fewer output tokens while maintaining high performance.
- It supports text and images with a 1-million-token context window, runs locally on specific hardware like Macs with 32 GB of memory, and is available under an Apache 2.0 license.
Why it matters
Inkling 2 Mini advances local AI capabilities by achieving high benchmark performance with significantly fewer active parameters and lower computational costs, expanding efficient deployment options for developers.