Agens Volundr 32B v1: we fixed what you flagged in the Preview. 0/50 loops, session continuation, and an MIT engine that runs it with 262K context (measured to 239K) within a 16 GB budget

Explain Like I'm Normal

Blockway released v1 of their Volundr 32B model, addressing major community complaints regarding infinite loops and slow prefix caching. The model now supports a massive 262K context window within a 16GB VRAM budget using a custom MIT-developed engine. It significantly improves efficiency on SWE-bench tasks and reduces follow-up turn latency by nearly 3x.