diff --git a/recipes/deepseek-v4-1-flash-764b-on-4x-cmp-170hx-pp4-shadow-kv-exl3-2bpw/recipe.yaml b/recipes/deepseek-v4-1-flash-764b-on-4x-cmp-170hx-pp4-shadow-kv-exl3-2bpw/recipe.yaml
index 1639f3a2c..49f43cc59 100644
--- a/recipes/deepseek-v4-1-flash-764b-on-4x-cmp-170hx-pp4-shadow-kv-exl3-2bpw/recipe.yaml
+++ b/recipes/deepseek-v4-1-flash-764b-on-4x-cmp-170hx-pp4-shadow-kv-exl3-2bpw/recipe.yaml
@@ -208,3 +208,7 @@ provenance:
     - 'https://github.com/kaka86mm/dsv41-flash-pp4-170hx/blob/950f4c003c074be9434735a37ec67466559909ec/serve/launch-pp4.sh'
     - 'https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash'
   notes: "Community ingest of kaka86mm's recipe repo @ 950f4c0 (repo owner-named by Casey Oct 6 as part of the 170HX focus wave; Apache-2.0). Launch argv + runtime env transcribed verbatim from serve/launch-pp4.sh @ 950f4c0 (v37, includes the 09-28 GLOO_SOCKET_IFNAME=lo fix). Measured numbers are the author's README 09-28 v37 table + bench/reports/2026-09-14-v13-512k.md + FINDINGS.md §7, logged claimed_by=submitter T0 (no run file) — the repo's own measurement-discipline rules (bench/README, nonce-at-head, unique-text cold prefill, warmup 2 rounds) are noted because four of its earlier numbers were retracted in FINDINGS §4. Model: 764B native vision (image-text-to-text pipeline tag on HF); V4.1-Flash is its own model generation — separate row + family deepseek-v4-1-flash, NOT a variant of V4-Flash-0731 (site already carries both families since the jj Blackwell recipe). Built-on credits per the repo's own acknowledgements: zebgop-ops/dsv41reap-pp (shadow KV overlay, Apache-2.0), vcruz305/vllm-exl3 (EXL3 vLLM plugin, AGPL-3.0), turboderp-org/exllamav3 (quantization + kernels), wtdcode/vllm-backport (sm80 sparse-MLA lineage; ampere/ sparse-attention dir copied from its tree)."
+max_context_len: 524288
+kv_pool_tokens: 5895670
+storage_requirements:
+  ram_gb: 373.0