A text-focused open-weight model from the GLM family with an exceptionally large context window of roughly one million tokens — meaning it can hold and reason over very long documents in a single pass. Details about its specific strengths and behavioral quirks are limited beyond its architecture lineage and context capacity. The 'exlr8' naming suggests an optimized or quantized variant, likely tuned for faster inference.