LiquidAI: LFM2.5-1.2B-Thinking

Liquid AI's LFM2.5 Thinking line is positioned around reasoning under strict device, memory and latency constraints. This 1.2B Thinking entry is useful as a pointer for compact reasoning experiments where local or edge execution matters more than maximum model size. Verify the current model release, memory footprint, supported runtimes, license and benchmark claims directly with Liquid AI.