diff --git a/tools/server/server.cpp b/tools/server/server.cpp
index b84c119579..72ad53550f 100644
--- a/tools/server/server.cpp
+++ b/tools/server/server.cpp
@@ -2385,6 +2385,10 @@ struct server_context {
         llama_batch_free(batch);
     }
 
+    int32_t n_ctx_slot() const {
+        return params_base.kv_unified ? n_ctx : n_ctx / params_base.n_parallel;
+    }
+
     bool load_model(const common_params & params) {
         SRV_INF("loading model '%s'\n", params.model.path.c_str());
 
@@ -2413,7 +2417,7 @@ struct server_context {
 
             params_dft.devices      = params_base.speculative.devices;
             params_dft.model        = params_base.speculative.model;
-            params_dft.n_ctx        = params_base.speculative.n_ctx == 0 ? slots.front().n_ctx : params_base.speculative.n_ctx;
+            params_dft.n_ctx        = params_base.speculative.n_ctx == 0 ? n_ctx_slot() : params_base.speculative.n_ctx;
             params_dft.n_gpu_layers = params_base.speculative.n_gpu_layers;
             params_dft.n_parallel   = 1;
             params_dft.cache_type_k = params_base.speculative.cache_type_k;
@@ -2501,8 +2505,6 @@ struct server_context {
     }
 
     void init() {
-        const int32_t n_ctx_slot = params_base.kv_unified ? n_ctx : n_ctx / params_base.n_parallel;
-
         SRV_INF("initializing slots, n_slots = %d\n", params_base.n_parallel);
 
         for (int i = 0; i < params_base.n_parallel; i++) {
@@ -2510,7 +2512,7 @@ struct server_context {
 
             slot.id = i;
             slot.ctx = ctx;
-            slot.n_ctx = n_ctx_slot;
+            slot.n_ctx = n_ctx_slot();
             slot.mctx = mctx;
             slot.prompt.tokens.has_mtmd = mctx != nullptr;
 
@@ -2533,7 +2535,7 @@ struct server_context {
                 }
             }
 
-            SLT_INF(slot, "new slot n_ctx_slot = %d\n", slot.n_ctx);
+            SLT_INF(slot, "new slot, n_ctx = %d\n", slot.n_ctx);
 
             slot.callback_on_release = [this](int) {
                 queue_tasks.pop_deferred_task();
@@ -3718,7 +3720,7 @@ struct server_context {
                         slot.n_past = 0;
                         slot.state = SLOT_STATE_PROCESSING_PROMPT;
 
-                        SLT_INF(slot, "new prompt, n_ctx_slot = %d, n_keep = %d, n_prompt_tokens = %d\n",
+                        SLT_INF(slot, "new prompt, n_ctx = %d, n_keep = %d, n_prompt_tokens = %d\n",
                                 slot.n_ctx, slot.task->params.n_keep, slot.n_prompt_tokens());
 
                         // print prompt tokens (for debugging)