diff --git a/src/ailego/math/fht_dispatch.cc b/src/ailego/math/fht_dispatch.cc
index b1e9c4aac..9a02f3f04 100644
--- a/src/ailego/math/fht_dispatch.cc
+++ b/src/ailego/math/fht_dispatch.cc
@@ -46,6 +46,12 @@ void fht_flip_sign_neon(const uint8_t *flip, float *data, size_t dim);
 void fht_kacs_walk_neon(float *data, size_t len);
 void fht_inv_kacs_walk_neon(float *data, size_t len);
 #endif
+#if defined(__riscv_vector)
+void fht_flip_sign_rvv(const uint8_t *flip, float *data, size_t dim);
+void fht_kacs_walk_rvv(float *data, size_t len);
+void fht_inv_kacs_walk_rvv(float *data, size_t len);
+void fht_inplace_rvv(float *data, size_t n);
+#endif
 
 // ============================================================================
 // Runtime dispatch entry points
@@ -54,6 +60,8 @@ void fht_inv_kacs_walk_neon(float *data, size_t len);
 void fht_flip_sign(const uint8_t *flip, float *data, size_t dim) {
 #if defined(__ARM_NEON) && defined(__aarch64__)
   fht_flip_sign_neon(flip, data, dim);
+#elif defined(__riscv_vector)
+  fht_flip_sign_rvv(flip, data, dim);
 #else
 #if defined(__AVX512F__)
   if (internal::CpuFeatures::static_flags_.AVX512F &&
@@ -81,6 +89,8 @@ void fht_flip_sign(const uint8_t *flip, float *data, size_t dim) {
 void fht_kacs_walk(float *data, size_t len) {
 #if defined(__ARM_NEON) && defined(__aarch64__)
   fht_kacs_walk_neon(data, len);
+#elif defined(__riscv_vector)
+  fht_kacs_walk_rvv(data, len);
 #else
 #if defined(__AVX512F__)
   if (internal::CpuFeatures::static_flags_.AVX512F) {
@@ -107,6 +117,8 @@ void fht_kacs_walk(float *data, size_t len) {
 void fht_inv_kacs_walk(float *data, size_t len) {
 #if defined(__ARM_NEON) && defined(__aarch64__)
   fht_inv_kacs_walk_neon(data, len);
+#elif defined(__riscv_vector)
+  fht_inv_kacs_walk_rvv(data, len);
 #else
 #if defined(__AVX512F__)
   if (internal::CpuFeatures::static_flags_.AVX512F) {
@@ -131,6 +143,9 @@ void fht_inv_kacs_walk(float *data, size_t len) {
 }
 
 void fht_inplace(float *data, size_t n) {
+#if defined(__riscv_vector)
+  fht_inplace_rvv(data, n);
+#else
 #if defined(__AVX512F__)
   if (internal::CpuFeatures::static_flags_.AVX512F) {
     fht_inplace_avx512(data, n);
@@ -144,6 +159,7 @@ void fht_inplace(float *data, size_t n) {
   }
 #endif
   fht_inplace_scalar(data, n);
+#endif  // __riscv_vector
 }
 
 }  // namespace ailego
diff --git a/src/ailego/math/fht_rvv.cc b/src/ailego/math/fht_rvv.cc
new file mode 100644
index 000000000..bc5a37bb1
--- /dev/null
+++ b/src/ailego/math/fht_rvv.cc
@@ -0,0 +1,102 @@
+// Copyright 2025-present the zvec project
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+#if defined(__riscv_vector)
+
+#include <cmath>
+#include <cstddef>
+#include <cstdint>
+#include <riscv_vector.h>
+
+namespace zvec {
+namespace ailego {
+
+void fht_flip_sign_rvv(const uint8_t *flip, float *data, size_t dim) {
+  const uint32_t sign_bit = 0x80000000u;
+  size_t i = 0;
+  while (i < dim) {
+    size_t vl = __riscv_vsetvl_e32m8(dim - i);
+    vbool4_t mask = __riscv_vlm_v_b4(flip + i / 8, vl);
+    vfloat32m8_t v = __riscv_vle32_v_f32m8(data + i, vl);
+    vuint32m8_t bits = __riscv_vreinterpret_v_f32m8_u32m8(v);
+    bits = __riscv_vxor_vx_u32m8_mu(mask, bits, bits, sign_bit, vl);
+    __riscv_vse32_v_f32m8(data + i, __riscv_vreinterpret_v_u32m8_f32m8(bits),
+                          vl);
+    i += vl;
+  }
+}
+
+void fht_kacs_walk_rvv(float *data, size_t len) {
+  size_t half = len / 2;
+  size_t base = len % 2;
+  size_t offset = base + half;
+  size_t i = 0;
+  while (i < half) {
+    size_t vl = __riscv_vsetvl_e32m8(half - i);
+    vfloat32m8_t x = __riscv_vle32_v_f32m8(data + i, vl);
+    vfloat32m8_t y = __riscv_vle32_v_f32m8(data + i + offset, vl);
+    __riscv_vse32_v_f32m8(data + i, __riscv_vfadd_vv_f32m8(x, y, vl), vl);
+    __riscv_vse32_v_f32m8(data + i + offset, __riscv_vfsub_vv_f32m8(x, y, vl),
+                          vl);
+    i += vl;
+  }
+  if (base != 0) {
+    data[half] *= std::sqrt(2.0f);
+  }
+}
+
+void fht_inv_kacs_walk_rvv(float *data, size_t len) {
+  size_t half = len / 2;
+  size_t base = len % 2;
+  size_t offset = base + half;
+  if (base != 0) {
+    data[half] *= std::sqrt(0.5f);
+  }
+  size_t i = 0;
+  while (i < half) {
+    size_t vl = __riscv_vsetvl_e32m8(half - i);
+    vfloat32m8_t a = __riscv_vle32_v_f32m8(data + i, vl);
+    vfloat32m8_t b = __riscv_vle32_v_f32m8(data + i + offset, vl);
+    vfloat32m8_t t = __riscv_vfadd_vv_f32m8(a, b, vl);
+    __riscv_vse32_v_f32m8(data + i, __riscv_vfmul_vf_f32m8(t, 0.5f, vl), vl);
+    t = __riscv_vfsub_vv_f32m8(a, b, vl);
+    __riscv_vse32_v_f32m8(data + i + offset,
+                          __riscv_vfmul_vf_f32m8(t, 0.5f, vl), vl);
+    i += vl;
+  }
+}
+
+void fht_inplace_rvv(float *data, size_t n) {
+  for (size_t len = 1; len < n; len <<= 1) {
+    size_t step = len << 1;
+    for (size_t i = 0; i < n; i += step) {
+      size_t j = 0;
+      while (j < len) {
+        size_t vl = __riscv_vsetvl_e32m8(len - j);
+        vfloat32m8_t u = __riscv_vle32_v_f32m8(data + i + j, vl);
+        vfloat32m8_t v = __riscv_vle32_v_f32m8(data + i + j + len, vl);
+        __riscv_vse32_v_f32m8(data + i + j, __riscv_vfadd_vv_f32m8(u, v, vl),
+                              vl);
+        __riscv_vse32_v_f32m8(data + i + j + len,
+                              __riscv_vfsub_vv_f32m8(u, v, vl), vl);
+        j += vl;
+      }
+    }
+  }
+}
+
+}  // namespace ailego
+}  // namespace zvec
+
+#endif  // __riscv_vector