diff --git a/ggml/include/ggml.h b/ggml/include/ggml.h
index 888f00c2e8..48da68fe7e 100644
--- a/ggml/include/ggml.h
+++ b/ggml/include/ggml.h
@@ -539,7 +539,6 @@ extern "C" {
         GGML_OP_FLASH_ATTN_BACK,
         GGML_OP_SSM_CONV,
         GGML_OP_SSM_SCAN,
-        GGML_OP_KDA_SCAN,
         GGML_OP_WIN_PART,
         GGML_OP_WIN_UNPART,
         GGML_OP_GET_REL_POS,
@@ -2338,28 +2337,6 @@ extern "C" {
             struct ggml_tensor  * C,
             struct ggml_tensor  * ids);
 
-    // KDA (Kimi Delta Attention) scan
-    // Delta attention recurrence:
-    //   h[t] = exp(g[t]) * h[t-1] + k[t]^T * (beta[t] * (v[t] - h[t-1] @ k[t]))
-    //   o[t] = q[t]^T @ h[t]
-    // Parameters:
-    //   h:    hidden state {head_dim, head_dim, n_head, n_seqs+}
-    //   q:    query        {head_dim, n_head, n_seq_tokens, n_seqs}
-    //   k:    key          {head_dim, n_head, n_seq_tokens, n_seqs}
-    //   v:    value        {head_dim, n_head, n_seq_tokens, n_seqs}
-    //   g:    gate         {head_dim, n_head, n_seq_tokens, n_seqs}
-    //   beta: mixing       {n_head, n_seq_tokens, n_seqs}
-    //   ids:  seq indices  {n_seqs}
-    GGML_API struct ggml_tensor * ggml_kda_scan(
-            struct ggml_context * ctx,
-            struct ggml_tensor  * h,
-            struct ggml_tensor  * q,
-            struct ggml_tensor  * k,
-            struct ggml_tensor  * v,
-            struct ggml_tensor  * g,
-            struct ggml_tensor  * beta,
-            struct ggml_tensor  * ids);
-
     // partition into non-overlapping windows with padding if needed
     // example:
     // a:   768   64   64    1
diff --git a/ggml/src/ggml-cpu/ggml-cpu.c b/ggml/src/ggml-cpu/ggml-cpu.c
index 4cc15b0981..3247af8bb0 100644
--- a/ggml/src/ggml-cpu/ggml-cpu.c
+++ b/ggml/src/ggml-cpu/ggml-cpu.c
@@ -2320,7 +2320,6 @@ static int ggml_get_n_tasks(struct ggml_tensor * node, int n_threads) {
         case GGML_OP_FLASH_ATTN_BACK:
         case GGML_OP_SSM_CONV:
         case GGML_OP_SSM_SCAN:
-        case GGML_OP_KDA_SCAN:
         case GGML_OP_RWKV_WKV6:
         case GGML_OP_GATED_LINEAR_ATTN:
         case GGML_OP_RWKV_WKV7:
diff --git a/ggml/src/ggml-cuda/ggml-cuda.cu b/ggml/src/ggml-cuda/ggml-cuda.cu
index bbd7810dd2..eb2e273110 100644
--- a/ggml/src/ggml-cuda/ggml-cuda.cu
+++ b/ggml/src/ggml-cuda/ggml-cuda.cu
@@ -4503,11 +4503,6 @@ static bool ggml_backend_cuda_device_supports_op(ggml_backend_dev_t dev, const g
                 return op->src[0]->ne[0] == 16 && op->src[0]->ne[1] == 1 && op->src[0]->ne[2] % 128 == 0 && op->src[4]->ne[1] == 1;
             }
         }
-        case GGML_OP_KDA_SCAN: {
-            // KDA scan kernel supports head_dim 64 or 128
-            const int64_t head_dim = op->src[0]->ne[0];
-            return head_dim == 64 || head_dim == 128;
-        }
         case GGML_OP_SSM_CONV: {
             // assumes d_inner % threads == 0
             return op->src[0]->ne[1] % 128 == 0;
diff --git a/ggml/src/ggml.c b/ggml/src/ggml.c
index a167d6a574..173ec6b98f 100644
--- a/ggml/src/ggml.c
+++ b/ggml/src/ggml.c
@@ -999,7 +999,6 @@ static const char * GGML_OP_NAME[GGML_OP_COUNT] = {
     "FLASH_ATTN_BACK",
     "SSM_CONV",
     "SSM_SCAN",
-    "KDA_SCAN",
     "WIN_PART",
     "WIN_UNPART",
     "GET_REL_POS",