diff --git a/recipes/glm-5-3-flash-4x6000-kk-mtp3/recipe.yaml b/recipes/glm-5-3-flash-4x6000-kk-mtp3/recipe.yaml
index ac67ddf69..f26ef8300 100644
--- a/recipes/glm-5-3-flash-4x6000-kk-mtp3/recipe.yaml
+++ b/recipes/glm-5-3-flash-4x6000-kk-mtp3/recipe.yaml
@@ -143,3 +143,4 @@ provenance:
     - 'https://github.com/local-inference-lab/rtx6kpro/blob/master/benchmarks/karmic-kraken-serving.md'
     - 'https://github.com/local-inference-lab/rtx6kpro/blob/master/benchmarks/karmic-kraken-serving-samples.json'
   notes: "Benchmark-qualified configuration from local-inference-lab/rtx6kpro (Karmic Kraken serving report). Five warmed 30-second windows per cell on RTX PRO 6000 Blackwell Max-Q; decode ctx-0 temp 1. Launch argv transcribed verbatim from the repo's unified Docker guide (same composition). Measured: C1 281.3 tok/s, C8 897.6 tok/s, 32K prefill 13816 tok/s."
+max_context_len: 1048576