diff --git a/cpp/src/neighbors/detail/cagra/cagra_build.cuh b/cpp/src/neighbors/detail/cagra/cagra_build.cuh
index a7c15b4161..f118146fba 100644
--- a/cpp/src/neighbors/detail/cagra/cagra_build.cuh
+++ b/cpp/src/neighbors/detail/cagra/cagra_build.cuh
@@ -1,11 +1,12 @@
 /*
- * SPDX-FileCopyrightText: Copyright (c) 2023-2026, NVIDIA CORPORATION.
- * SPDX-License-Identifier: Apache-2.0
+ * SPDX-FileCopyrightText: Copyright (c) 2023-2026, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+ * reserved. SPDX-License-Identifier: Apache-2.0
  */
 #pragma once
 
 #include "../../../core/nvtx.hpp"
 #include "../../../preprocessing/quantize/vpq_build-ext.cuh"
+#include "../../ivf_pq/ivf_pq_fp16_overflow.cuh"
 #include "graph_core.cuh"
 
 #include <raft/core/copy.cuh>
@@ -1664,6 +1665,24 @@ void build_knn_graph(
   RAFT_LOG_DEBUG("# Building IVF-PQ index %s", model_name.c_str());
   auto index = cuvs::neighbors::ivf_pq::build(res, pq.build_params, dataset);
 
+  // Empirically detect FP16 distance overflow on the just-built index: run a small FP16 probe
+  // search and downgrade the internal/coarse dtypes to FP32 if any distance comes back non-finite.
+  // This observes the actual computation, so it is agnostic of the selected distance type.
+  if (pq.search_params.internal_distance_dtype == CUDA_R_16F ||
+      pq.search_params.coarse_search_dtype == CUDA_R_16F) {
+    {
+      const bool fp16_overflow = cuvs::neighbors::ivf_pq::helpers::detect_fp16_overflow(
+        res, index, pq.search_params, dataset);
+      if (fp16_overflow) {
+        RAFT_LOG_WARN(
+          "IVF-PQ FP16 distance produced non-finite results on a probe search for this dataset -> "
+          "switching 'internal_distance_dtype' and 'coarse_search_dtype' to FP32");
+        pq.search_params.internal_distance_dtype = CUDA_R_32F;
+        pq.search_params.coarse_search_dtype     = CUDA_R_32F;
+      }
+    }
+  }
+
   //
   // search top (k + 1) neighbors
   //
@@ -2217,6 +2236,7 @@ index<T, IdxT> build(
       knn_build_params = cagra::graph_build_params::ivf_pq_params(dataset.extents(), params.metric);
     }
   }
+
   RAFT_EXPECTS(
     params.metric != cuvs::distance::DistanceType::BitwiseHamming ||
       std::holds_alternative<cagra::graph_build_params::iterative_search_params>(
diff --git a/cpp/src/neighbors/ivf_pq/ivf_pq_fp16_overflow.cuh b/cpp/src/neighbors/ivf_pq/ivf_pq_fp16_overflow.cuh
new file mode 100644
index 0000000000..4cd239af5d
--- /dev/null
+++ b/cpp/src/neighbors/ivf_pq/ivf_pq_fp16_overflow.cuh
@@ -0,0 +1,84 @@
+/*
+ * SPDX-FileCopyrightText: Copyright (c) 2026, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+ * SPDX-License-Identifier: Apache-2.0
+ */
+
+#pragma once
+
+#include "../detail/ann_utils.cuh"  // cuvs::spatial::knn::detail::utils::mapping
+
+#include <cuvs/distance/distance.hpp>
+#include <cuvs/neighbors/ivf_pq.hpp>
+
+#include <raft/core/device_mdarray.hpp>
+#include <raft/core/error.hpp>
+#include <raft/core/mdspan.hpp>
+#include <raft/core/operators.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
+#include <raft/core/resource/device_memory_resource.hpp>
+#include <raft/core/resource/thrust_policy.hpp>
+#include <raft/core/resources.hpp>
+#include <raft/linalg/map_reduce.cuh>
+#include <raft/linalg/reduce.cuh>
+#include <raft/util/cuda_dev_essentials.cuh>
+#include <raft/util/cudart_utils.hpp>
+
+#include <thrust/logical.h>
+
+#include <algorithm>
+#include <cstdint>
+
+namespace cuvs::neighbors::ivf_pq::helpers {
+
+/**
+ * @brief Detect whether FP16 internal distance dtypes overflow for this dataset during search.
+ *
+ * Runs a small probe search against an already-built IVF-PQ index with current distance types,
+ * and reports whether any returned distance is non-finite (inf/NaN).
+ */
+template <typename DataT, typename Accessor>
+bool detect_fp16_overflow(
+  raft::resources const& handle,
+  const cuvs::neighbors::ivf_pq::index<int64_t>& index,
+  cuvs::neighbors::ivf_pq::search_params search_params,
+  raft::mdspan<const DataT, raft::matrix_extent<int64_t>, raft::row_major, Accessor> dataset)
+{
+  const int64_t n_rows = dataset.extent(0);
+  if (n_rows == 0) { return false; }
+
+  auto stream       = raft::resource::get_cuda_stream(handle);
+  const int64_t dim = dataset.extent(1);
+
+  constexpr int64_t kMaxSampleQueries = 128;
+  constexpr uint32_t kProbeTopK       = 32;
+  const int64_t n_sample              = std::min<int64_t>(n_rows, kMaxSampleQueries);
+  const uint32_t top_k = std::min<uint32_t>(static_cast<uint32_t>(n_rows), kProbeTopK);
+
+  auto mr = raft::resource::get_workspace_resource_ref(handle);
+  auto queries =
+    raft::make_device_mdarray<DataT>(handle, mr, raft::make_extents<int64_t>(n_sample, dim));
+  raft::copy(queries.data_handle(), dataset.data_handle(), n_sample * dim, stream);
+
+  auto neighbors =
+    raft::make_device_mdarray<int64_t>(handle, mr, raft::make_extents<int64_t>(n_sample, top_k));
+  auto distances =
+    raft::make_device_mdarray<float>(handle, mr, raft::make_extents<int64_t>(n_sample, top_k));
+
+  cuvs::neighbors::ivf_pq::search(handle,
+                                  search_params,
+                                  index,
+                                  raft::make_const_mdspan(queries.view()),
+                                  neighbors.view(),
+                                  distances.view());
+
+  const int64_t count       = n_sample * static_cast<int64_t>(top_k);
+  auto is_non_finite_op     = [] __device__(float v) { return isnan(v) || isinf(v); };
+  const bool any_non_finite = thrust::any_of(raft::resource::get_thrust_policy(handle),
+                                             distances.data_handle(),
+                                             distances.data_handle() + count,
+                                             is_non_finite_op);
+  raft::resource::sync_stream(handle);
+  return any_non_finite;
+}
+
+}  // namespace cuvs::neighbors::ivf_pq::helpers