diff --git a/llamacpp/README.md b/llamacpp/README.md
index fcffee2..73a5af9 100644
--- a/llamacpp/README.md
+++ b/llamacpp/README.md
@@ -52,24 +52,25 @@ If only a CPU device shows up, the container can't see the GPU — check that
 
 Both models run simultaneously on the same 96 GiB VRAM pool. Approximate usage:
 
-| Model             | Weights  | KV cache (32k×4) | Subtotal |
-|-------------------|----------|-------------------|----------|
-| qwen3.6-27b       | ~16 GiB  | ~34 GiB           | ~50 GiB  |
-| qwen3.6-35b-a3b   | ~20 GiB  | ~10 GiB           | ~30 GiB  |
-| **Total**         |          |                   | **~80 GiB** |
+| Model             | Weights  | KV cache (131k×4 for a3b, 32k×4 for 27b) | Subtotal |
+|-------------------|----------|-------------------------------------------|----------|
+| qwen3.6-27b       | ~16 GiB  | ~8 GiB  (32k total, 8k/slot)              | ~24 GiB  |
+| qwen3.6-35b-a3b   | ~20 GiB  | ~9 GiB  (131k total, 33k/slot)            | ~29 GiB  |
+| **Total**         |          |                                           | **~53 GiB** |
 
-~16 GiB headroom — comfortable but not infinite. If VRAM is exhausted ( Vulkan
-allocation failures in logs), reduce `-c` on the 27B (its KV cache dominates) or
-drop `-np` to 2 on either model.
+~43 GiB headroom — very comfortable. The MoE's KV cache is tiny (~72 KiB/token
+vs the dense 27B's ~256 KiB/token), so large context is nearly free.
 
 ## Tuning
 
 The key knobs (in each `deployment-*.yaml`):
 
 - `-ngl 999` — offload all layers to GPU. Reduce only if VRAM is tight.
-- `-c 32768` — total KV-cache context. With `-np 4` this is 8192 tokens per
-  concurrent request. The 27B's dense KV cache is the larger consumer (~34 GiB
-  at 32k×4); the MoE's is much smaller (~10 GiB).
+- `-c 131072` (35b-a3b) / `32768` (27b) — total KV-cache context. With `-np 4`
+  the 35b-a3b gets 32768 tokens per slot (enough for the full SOUL.md + prompt);
+  the 27b gets 8192 per slot. The MoE's KV cache is ~72 KiB/token so large
+  context is cheap; the dense 27B's is ~256 KiB/token. Raise the 27B's `-c` too
+  if interactive sessions hit the context limit.
 - `-np 4` — parallel slots (concurrent requests). Each extra slot multiplies
   KV-cache VRAM usage. Bump higher on the flash model if you need more
   throughput (it has VRAM headroom).
diff --git a/llamacpp/deployment-qwen36-35b-a3b.yaml b/llamacpp/deployment-qwen36-35b-a3b.yaml
index 28af4d8..77cd695 100644
--- a/llamacpp/deployment-qwen36-35b-a3b.yaml
+++ b/llamacpp/deployment-qwen36-35b-a3b.yaml
@@ -85,8 +85,8 @@ spec:
         - -ngl                     # offload ALL layers to the GPU (fits in 96 GiB VRAM)
         - "999"
         - -c                       # total KV-cache context, split across parallel slots
-        - "32768"
-        - -np                      # 4 parallel slots => 8192 tokens per concurrent request
+        - "131072"
+        - -np                      # 4 parallel slots => 32768 tokens per concurrent request
         - "4"
         - --cont-batching          # continuous batching across slots
         - --threads                # CPU threads for sampling/overhead (GPU does the heavy lifting)